Totally-enclosed dual-cooling waterproof pump self-adaptive control system and method

Abstract

The present application belongs to the field of water pump regulation and control, and particularly relates to a fully-sealed double-cooling water pump self-adaptive control system and method, which comprises: a sensing module that collects sensing signals in real time from sensors disposed at key positions of the water pump; an adjustment module that calculates the motor input power and total heat load based on the sensing signals, respectively calculates the current cooling contribution of the internal shaft cooling loop and the external forced air cooling loop, and distributes the target cooling contribution of the two loops and converts it into an adjustment instruction for execution under the constraint that the temperature rise at the temperature measurement point does not exceed the safety threshold; a water leakage early warning module that outputs a leakage risk level and triggers a graded early warning through a prediction model based on the leakage signal, humidity and operation time; and an online cleaning module that determines the degree of pollution based on the temperature difference, the attenuation amplitude of the heat dissipation efficiency and the cumulative operation time, and automatically starts a cleaning program; the present application realizes the collaborative self-adaptive adjustment of the double-cooling system, the graded early warning of the sealing state and the intelligent control of the online cleaning.

Description

Technical Field

[0001] This invention belongs to the field of water pump regulation, and particularly relates to an adaptive control system and method for a fully sealed, dual-cooled, waterproof pump. Background Technology

[0002] In the operation of a fully sealed waterproof pump integrating in-shaft self-circulation cooling and external forced air cooling, the working characteristics of its dual cooling systems differ fundamentally: in-shaft self-circulation cooling utilizes the transported medium itself as coolant, and its cooling capacity is limited by the medium temperature, flow rate, and loop unobstructedness; external forced air cooling relies on cooling fan blades driving airflow through the heat dissipation fins to remove heat, and its cooling effect is affected by ambient temperature, fin surface cleanliness, and fan speed. However, under the existing control method, both cooling systems operate independently with fixed parameters, unable to perceive their respective actual cooling contribution rates, nor can they be coordinated and allocated according to real-time heat load. When the transported medium temperature is low, the ambient temperature is moderate, and the motor is lightly loaded, the dual cooling system still operates at full load, resulting in ineffective energy consumption; when the medium temperature rises or the motor is heavily loaded, causing the total heat load to exceed the capacity of a single cooling loop, the system cannot automatically increase the compensating cooling capacity of the other loop, causing excessive local temperature rise in the motor windings or bearings, accelerating insulation aging and lubrication failure.

[0003] While a fully sealed structure isolates external liquids from intrusion at the hardware level, it also makes the internal condition invisible: whether the medium circulation in the internal cooling circuit is smooth, whether the flow resistance on the inner wall of the circuit increases due to scaling, and whether there are minor leaks in the seals after long-term operation cannot be directly detected during operation. Current methods can only be used for shutdown inspection or tracing after a fault, lacking the ability to continuously sense and predict trends in the service state. Similarly, dust accumulation and blockage of external heat sink fins will reduce air cooling efficiency. The deterioration process is slow and difficult to detect. By the time the motor temperature becomes abnormal, it is often severely clogged. At this point, post-incident cleaning is insufficient to restore the original heat dissipation performance, and the timing of cleaning relies entirely on manual experience.

[0004] In existing technologies, some solutions use temperature sensors installed at the motor windings or bearings for over-temperature alarms, but this only achieves passive protection with a single threshold and cannot drive dynamic matching and adjustment of the dual cooling system. Other solutions introduce humidity sensors into the motor cavity to detect water ingress, but humidity changes caused by condensation and leakage are difficult to distinguish in a single sensor signal, leading to both false alarms and missed alarms, and failing to provide graded early warning of the aging and degradation process of the seals. Regarding online cleaning, existing solutions use timed automatic cleaning devices, but the timing strategy cannot be correlated with the actual degree of contamination. Over-cleaning in lightly contaminated conditions wastes water and cleaning media, while excessively long cleaning intervals in heavily contaminated conditions cause a continuous decline in heat dissipation capacity. Furthermore, the various sensor signals and execution control are isolated from each other, failing to form a closed-loop information link from heat load perception, cooling capacity assessment, seal status inference to cleaning decision generation. When an anomaly occurs in any link, the system cannot trace the root cause of the anomaly, nor can it provide preventative maintenance instructions in the early stages of fault evolution.

[0005] In summary, the primary issue is how to accurately and in real time perceive the actual cooling efficiency of each of the two cooling systems under the condition that the internal state of the fully sealed structure cannot be directly observed, and how to achieve adaptive and coordinated allocation between the two based on the overall heat load changes, so as to avoid local overheating caused by insufficient cooling and energy waste caused by excessive cooling. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes an adaptive control system and method for a fully sealed dual-cooling waterproof pump. The sensing module collects real-time sensor signals including the inlet and outlet temperatures and flow rates of the internal cooling circuit, the temperature of the motor stator windings and front and rear bearings, the temperature of the casing and heat sink fins, leakage signals at the sealing surfaces, internal humidity of the motor, ambient temperature at the air inlet, motor current and voltage, and pump speed. The adjustment module calculates the motor input power and total heat load based on these sensor signals, and calculates the current cooling contribution of the internal cooling circuit and the external forced air cooling circuit respectively. With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise at the measuring point does not exceed a safety threshold, the module dynamically allocates the target cooling contribution of the two circuits and converts it into adjustment commands for execution. The waterproof warning module outputs the leakage risk level and triggers graded warnings based on leakage signals, humidity, and operating time using a predictive model. The online cleaning module determines the degree of contamination based on temperature difference, heat dissipation efficiency attenuation, and cumulative operating time, and automatically initiates the cleaning program. This invention achieves coordinated adaptive adjustment of the dual cooling system, graded warnings for sealing status, and intelligent online cleaning control, improving the pump's operating efficiency and reliability.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A fully sealed dual-cooled waterproof pump adaptive control system, wherein the fully sealed dual-cooled waterproof pump integrates an internal shaft cooling circuit and an external forced air cooling circuit, the system comprising:

[0009] The sensing module is configured to: collect sensor signals deployed at key locations on the water pump in real time;

[0010] The adjustment module is configured to: calculate the motor input power based on the sensor signal using the motor current and motor voltage, and determine the total heat load using the motor stator winding temperature, front bearing temperature, rear bearing temperature, and motor input power;

[0011] Meanwhile, the current cooling contribution of the internal cooling circuit is calculated from the difference between the inlet and outlet temperatures of the internal cooling circuit and the flow rate of the internal cooling circuit. The current cooling contribution of the external forced air cooling circuit is calculated from the difference between the ambient temperature at the far end of the air inlet and the temperature of the heat dissipation fins and the speed of the cooling fan blades.

[0012] With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated. At the same time, they are converted into corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands, and the corresponding in-shaft cooling flow rate adjustment commands and / or cooling fan speed adjustment commands are executed.

[0013] Specifically, it also includes:

[0014] The waterproof early warning module is configured to: input the leakage signal of the sealing joint surface, the internal humidity of the motor and the running time into the pre-constructed sealing failure prediction model based on the sensor signal, output the evaluation result characterizing the aging degree of the seal and the leakage risk level, and trigger the corresponding graded early warning signal according to the leakage risk level.

[0015] The online cleaning module is configured to: calculate a first temperature difference based on the sensor signal, using the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet; calculate a second temperature difference based on the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet; and calculate the current actual heat dissipation efficiency based on the second temperature difference and the current cooling contribution of the external forced air cooling circuit.

[0016] The current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency corresponding to the clean state under the same working conditions to obtain the heat dissipation efficiency attenuation range. Combining the first temperature difference, the heat dissipation efficiency attenuation range and the cumulative running time, the degree of external pollution of the water pump is determined. When the degree of pollution reaches the preset cleaning trigger threshold, the online cleaning program is started to control the start and stop of the cleaning device, the cleaning water pressure and the cleaning time.

[0017] Specifically, determining the total heat load includes:

[0018] The pump shaft power under the current operating condition is determined by the pump speed and the pre-determined pump head-speed-power mapping relationship, and the difference between the motor input power and the pump shaft power is taken as the current total power loss of the motor.

[0019] The stator copper loss component in the current total power loss is corrected based on the temperature coefficient of the copper resistance of the motor stator winding, and the copper loss correction amount is obtained. The stator copper loss component is calculated from the motor current and the room temperature resistance of the stator winding.

[0020] The front bearing temperature is compared with the normal temperature reference value of the front bearing under the current speed and load conditions to obtain the front bearing temperature rise deviation.

[0021] Specifically, determining the total heat load also includes:

[0022] The rear bearing temperature is compared with the normal temperature reference value of the rear bearing under the current speed and load conditions to obtain the rear bearing temperature rise deviation.

[0023] The temperature rise deviation of the front bearing and the temperature rise deviation of the rear bearing are input into a pre-established bearing friction loss increment mapping relationship to obtain the bearing friction loss increment.

[0024] The sum of the copper loss correction and the bearing friction loss increment is used as the loss correction, which is then added to the current total loss power to obtain the total heat load.

[0025] Specifically, the current cooling contribution of the internal cooling circuit is calculated, including:

[0026] The average value of the inlet temperature and the outlet temperature is taken as the qualitative temperature of the cooling medium. Based on the qualitative temperature, the specific heat capacity and density of the cooling medium at the current temperature are obtained from the cooling medium property parameter mapping table preset in the adjustment module.

[0027] At the same time, it is determined whether the flow rate of the cooling circuit inside the shaft is lower than the preset low flow rate threshold. If it is lower than the low flow rate threshold, the flow rate is corrected by a preset low flow rate correction coefficient, and the corrected flow rate is used as the flow rate value to be calculated. Otherwise, the collected flow rate value is used directly.

[0028] Specifically, calculating the current cooling contribution of the internal cooling circuit also includes:

[0029] The product of the flow rate value, density, specific heat capacity, and the difference between the outlet temperature and the inlet temperature is taken as the theoretical cooling contribution of the internal cooling circuit.

[0030] The estimated heat flow rate transferred by the motor side cooling medium is obtained by calculating the difference between the motor stator winding temperature, front bearing temperature, and rear bearing temperature and the inlet temperature, multiplying each difference by its respective preset equivalent heat transfer coefficient, and then summing them.

[0031] The ratio of the estimated heat flow to the theoretical cooling contribution is used as the thermal resistance correction coefficient under the current operating conditions. When the ratio is greater than 1, the theoretical cooling contribution is divided by the thermal resistance correction coefficient, and the result is the corrected current cooling contribution of the in-shaft cooling circuit.

[0032] Specifically, the current cooling contribution of the external forced air cooling circuit is calculated, including:

[0033] The cooling fan speed is obtained, and the current cooling fan speed is converted into the theoretical cooling air volume based on the mapping relationship between the fan speed and the theoretical air volume preset in the adjustment module.

[0034] The pressure difference value on both sides of the air inlet filter is obtained, and the pressure difference value is compared with the reference pressure difference value of the filter in the clean state preset in the adjustment module to obtain the filter blockage pressure difference increment; at the same time, the fin temperature values ​​collected by multiple preset temperature measuring points on the surface of the heat dissipation fins are obtained, and the arithmetic mean of the multiple fin temperature values ​​is taken as the fin average temperature.

[0035] Specifically, calculating the current cooling contribution of the external forced air cooling circuit also includes:

[0036] Based on the filter clogging pressure difference increment, the airflow attenuation coefficient under the current operating condition is obtained from the mapping relationship between the filter clogging pressure difference increment and the airflow attenuation coefficient preset in the adjustment module; the theoretical cooling airflow is multiplied by the airflow attenuation coefficient to obtain the effective cooling airflow.

[0037] The ambient temperature at the far end of the air inlet is obtained, and the difference between the average temperature of the fins and the ambient temperature at the far end of the air inlet is taken as the effective heat exchange temperature difference.

[0038] The current cooling contribution of the external forced air cooling circuit is calculated based on the effective cooling air volume, the effective heat exchange temperature difference, the specific heat capacity of air at constant pressure, and the air density.

[0039] Specifically, the target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated, including:

[0040] Obtain the total heat load, the current cooling contribution of the in-shaft cooling circuit, and the current cooling contribution of the external forced air cooling circuit; within a preset in-shaft cooling flow rate adjustment range, discretize into multiple candidate in-shaft cooling contributions according to the first step length; within a preset cooling fan speed adjustment range, discretize into multiple candidate external air cooling contributions according to the second step length; combine the multiple candidate in-shaft cooling contributions and the multiple candidate external air cooling contributions in pairs to form a candidate allocation scheme set;

[0041] For each candidate allocation scheme in the candidate allocation scheme set, based on the temperature rise response matrix preset in the adjustment module, the predicted temperature rise of the motor stator winding, the predicted temperature rise of the front bearing, and the predicted temperature rise of the rear bearing under the candidate allocation scheme are calculated respectively; the temperature rise response matrix represents the independent influence coefficient and cross influence coefficient of the unit shaft cooling contribution increment and the unit external air cooling contribution increment on the temperature rise of the motor stator winding, the front bearing, and the rear bearing.

[0042] Specifically, the target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit also include:

[0043] Remove any candidate allocation scheme whose predicted temperature rise exceeds its corresponding safety threshold from the candidate allocation scheme set, and retain the candidate allocation schemes in which the predicted temperature rise does not exceed the corresponding safety threshold to form a feasible allocation scheme set.

[0044] For each feasible allocation scheme in the set of feasible allocation schemes, based on the energy consumption function of the in-shaft cooling circuit and the energy consumption function of the external forced air cooling circuit preset in the adjustment module, the estimated energy consumption of the in-shaft cooling circuit and the estimated energy consumption of the external forced air cooling circuit under the corresponding feasible allocation scheme, as well as the estimated total cooling energy consumption of each feasible allocation scheme, are calculated respectively; the energy consumption function of the in-shaft cooling circuit is a function of the contribution of in-shaft cooling, and the energy consumption function of the external forced air cooling circuit is a function of the contribution of external air cooling.

[0045] Specifically, the target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit also include:

[0046] From the set of feasible allocation schemes, the feasible allocation scheme with the minimum estimated total cooling energy consumption is selected as the optimal allocation scheme. The candidate in-shaft cooling contribution in the optimal allocation scheme is taken as the target cooling contribution of the in-shaft cooling loop, and the candidate external air cooling contribution in the optimal allocation scheme is taken as the target cooling contribution of the external forced air cooling loop.

[0047] If the set of candidate allocation schemes is empty after elimination, then the boundary processing strategy is executed: the maximum output cooling contribution of the internal cooling circuit and the maximum output cooling contribution of the external forced air cooling circuit are respectively used as the target cooling contribution and output, and at the same time, the insufficient cooling capacity alarm signal is triggered.

[0048] An adaptive control method for a fully sealed, dual-cooled, waterproof pump includes:

[0049] Real-time acquisition of sensor signals deployed at key locations on the water pump;

[0050] Based on the sensor signals, the motor input power is calculated using the motor current and motor voltage, and the total heat load is determined using the motor stator winding temperature, front bearing temperature, rear bearing temperature, and motor input power.

[0051] Meanwhile, the current cooling contribution of the internal cooling circuit is calculated from the difference between the inlet and outlet temperatures of the internal cooling circuit and the flow rate of the internal cooling circuit. The current cooling contribution of the external forced air cooling circuit is calculated from the difference between the ambient temperature at the far end of the air inlet and the temperature of the heat dissipation fins and the speed of the cooling fan blades.

[0052] With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated. At the same time, they are converted into corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands, and the corresponding in-shaft cooling flow rate adjustment commands and / or cooling fan speed adjustment commands are executed.

[0053] Specifically, the adaptive control method for a fully sealed, dual-cooled, waterproof pump also includes:

[0054] Based on the sensor signals, the leakage signal of the sealing joint surface, the internal humidity of the motor and the running time are input into the pre-built sealing failure prediction model, and the assessment results characterizing the aging degree of the seal and the leakage risk level are output. The corresponding graded early warning signal is triggered according to the leakage risk level.

[0055] Based on the sensor signal, the first temperature difference is calculated from the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet, the second temperature difference is calculated from the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet, and the current actual heat dissipation efficiency is calculated from the second temperature difference and the current cooling contribution of the external forced air cooling circuit.

[0056] The current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency corresponding to the clean state under the same working conditions to obtain the heat dissipation efficiency attenuation range. Combining the first temperature difference, the heat dissipation efficiency attenuation range and the cumulative running time, the degree of external pollution of the water pump is determined. When the degree of pollution reaches the preset cleaning trigger threshold, the online cleaning program is started to control the start and stop of the cleaning device, the cleaning water pressure and the cleaning time.

[0057] Compared with the prior art, the beneficial effects of the present invention are:

[0058] This invention uses an adjustment module to calculate the motor input power and determine the total heat load in real time based on multi-parameter fusion sensing signals. It simultaneously calculates the current cooling contribution of the internal cooling circuit and the external forced air cooling circuit. With the minimum total cooling energy consumption as the optimization objective, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the two circuits is dynamically allocated and converted into corresponding adjustment commands for execution. This effectively solves the problem in the prior art that the dual cooling system operates independently with fixed parameters and cannot be coordinated and adjusted according to the real-time heat load. Specifically, based on the accurate determination of the total heat load, the adjustment module can accurately sense the current cooling capacity of each of the two circuits by obtaining the actual output cooling contribution of the internal cooling circuit and the external forced air cooling circuit respectively. On this basis, the allocation is optimized with the goal of minimizing the total cooling energy consumption, avoiding the ineffective energy consumption caused by the dual cooling system still operating at full load when the temperature of the conveying medium is low, the ambient temperature is moderate, and the motor is lightly loaded. At the same time, when the total heat load exceeds the capacity of a single circuit, the compensation cooling capacity of the other circuit is automatically increased to ensure that the temperature rise at each temperature measuring point is always controlled within the safe threshold. This achieves the adaptive and coordinated energy-saving adjustment of the dual cooling system and the synchronous improvement of motor operation reliability. Attached Figure Description

[0059] Figure 1 This is a flowchart of the fully sealed dual-cooled waterproof pump adaptive control system of Embodiment 1 of the present invention;

[0060] Figure 2 The flowchart for calculating the current cooling contribution of the external forced air cooling circuit is shown in Embodiment 2 of the present invention. Detailed Implementation

[0061] Example 1

[0062] Please see Figure 1 The present invention provides an embodiment of a fully sealed dual-cooling waterproof pump adaptive control system, wherein the fully sealed dual-cooling waterproof pump integrates an internal shaft cooling circuit and an external forced air cooling circuit, the internal shaft cooling circuit being specifically an internal shaft medium self-circulation cooling circuit; the system includes:

[0063] The sensing module is configured to: collect sensor signals deployed at key locations of the water pump in real time. The sensor signals include at least: inlet temperature of the internal cooling circuit, outlet temperature of the internal cooling circuit, flow rate of the internal cooling circuit, temperature of the motor stator winding, temperature of the front bearing, temperature of the rear bearing, surface temperature of the housing, temperature of the heat sink fins, leakage signal of the sealing joint surface, internal humidity of the motor, ambient temperature at the far end of the air inlet, motor current, motor voltage, water pump speed, water pump inlet pressure, and water pump outlet pressure.

[0064] The adjustment module is configured to: calculate the motor input power based on the sensor signal using the motor current and motor voltage, and determine the total heat load using the motor stator winding temperature, front bearing temperature, rear bearing temperature, and motor input power;

[0065] Meanwhile, the current cooling contribution of the internal cooling circuit is calculated from the difference between the inlet and outlet temperatures of the internal cooling circuit and the flow rate of the internal cooling circuit. The current cooling contribution of the external forced air cooling circuit is calculated from the difference between the ambient temperature at the far end of the air inlet and the temperature of the heat dissipation fins and the speed of the cooling fan blades.

[0066] With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated. At the same time, they are converted into corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands, and the corresponding in-shaft cooling flow rate adjustment commands and / or cooling fan speed adjustment commands are executed.

[0067] The waterproof early warning module is configured to: input the leakage signal of the sealing joint surface, the internal humidity of the motor and the running time into the pre-constructed sealing failure prediction model based on the sensor signal, output the evaluation result characterizing the aging degree of the seal and the leakage risk level, and trigger the corresponding graded early warning signal according to the leakage risk level.

[0068] The online cleaning module is configured to: calculate a first temperature difference based on the sensor signal, using the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet; calculate a second temperature difference based on the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet; and calculate the current actual heat dissipation efficiency based on the second temperature difference and the current cooling contribution of the external forced air cooling circuit.

[0069] The current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency corresponding to the clean state under the same working conditions to obtain the heat dissipation efficiency attenuation range. Combining the first temperature difference, the heat dissipation efficiency attenuation range and the cumulative running time, the degree of external pollution of the water pump is determined. When the degree of pollution reaches the preset cleaning trigger threshold, the online cleaning program is started to control the start and stop of the cleaning device, the cleaning water pressure and the cleaning time.

[0070] As a concrete example, consider a 110kW fully sealed, double-cooled, waterproof pump installed in the circulating water pump room of a chemical plant in southern coastal China. This pump is used to transport process cooling water with a temperature ranging from 55℃ to 70℃ and has been running continuously for approximately 800 hours. One afternoon, the sensor signals collected in real time by the sensing module showed: motor stator winding temperature 118℃, front bearing temperature 76℃, rear bearing temperature 73℃, inlet temperature of the internal cooling circuit 58℃, outlet temperature 62℃, internal cooling circuit flow rate 3.5 cubic meters per hour, casing surface temperature 52℃, average temperature of heat dissipation fins 49℃, leakage sensor output voltage at the sealing joint surface 0.1 volts, relative humidity inside the motor 62%, ambient temperature at the far end of the air inlet 33℃, motor current 156 amps, motor voltage 380 volts, pump speed 2850 rpm, cooling fan speed 1950 rpm, and pressure difference across the air inlet filter 95 Pascals. Based on the aforementioned sensor signals, the adjustment module calculated the motor input power to be 87.4kW, the total heat load to be 9.6kW, the current cooling contribution of the in-shaft cooling circuit to be 5.2kW, and the current cooling contribution of the external forced air cooling circuit to be 4.4kW. With the goal of minimizing total cooling energy consumption, and under the constraints of the motor stator winding temperature rise not exceeding 65℃, the front bearing temperature rise not exceeding 45℃, and the rear bearing temperature rise not exceeding 45℃, the optimized allocation yielded a target cooling contribution of 5.8kW for the in-shaft cooling circuit and 3.8kW for the external forced air cooling circuit. These were converted into adjustment commands to increase the in-shaft cooling flow rate from 3.5 cubic meters per hour to 4.1 cubic meters per hour and to reduce the cooling fan speed from 1950 rpm to 1720 rpm, respectively, and executed. After adjustment, the total cooling energy consumption decreased from 1.85kW to 1.62kW, a reduction of approximately 12.4%. Meanwhile, the waterproof early warning module inputs the 0.1V leakage signal from the sealing joint surface, the 62% internal humidity of the motor, and the cumulative operating time of 800 hours into the sealing failure prediction model. The model comprehensively judges that the leakage signal is within the noise floor range, the humidity change slope is only 1.2% per hour, and the dew point margin is 8℃. It outputs that the aging degree of the sealing component is mild aging and the leakage risk level is 0, so no early warning is triggered. The online cleaning module calculates that the equivalent thermal resistance of the casing increases by a factor of 1.15, the normalized heat dissipation efficiency decreases by 0.08, and the environmental weighted cumulative pollution time is 320 hours. The comprehensive pollution score is 18 points, and the pollution level is judged to be 0, i.e., clean state. The online cleaning program is not triggered, and the results of this assessment are only recorded in the local log.

[0071] Example 2

[0072] In this embodiment, since the heat sources such as stator copper loss, bearing friction loss, and rotor iron loss inside the motor are spatially dispersed and their heat transfer paths are coupled, the temperatures reflected by each temperature measurement point are not a direct mapping of a single heat source. In particular, the bearing temperature response to load changes has a significant lag. If the total heat load is directly calculated based solely on the stator winding temperature, bearing temperature, and motor input power, it is difficult to accurately distinguish the real-time contribution of each heat source under dynamic operating conditions, resulting in time lag and deviation in heat load estimation, which in turn affects the accuracy of the target cooling contribution allocation of the dual cooling circuit. To solve this problem, this embodiment provides a specific method for determining the total heat load, including:

[0073] A101. Calculate the motor input power based on the motor current and motor voltage, determine the pump shaft power under the current operating condition based on the pump speed and the pre-determined pump head-speed-power mapping relationship, and use the difference between the motor input power and the pump shaft power as the current total power loss of the motor.

[0074] It should be further explained that when calculating the current total power loss from the difference between the motor input power and the pump shaft power, it is necessary to obtain the key intermediate variable of pump shaft power in real time. However, pump shaft power cannot be directly measured by sensors during operation. In this embodiment, the pre-determined pump head-speed-power mapping relationship is determined during the pump's factory or installation and commissioning phase. Through bench testing, at least five speed operating points are selected at equal intervals within the pump's rated speed range. At each speed operating point, the pump outlet valve opening is adjusted to make the pump operate at the design flow point. Simultaneously, the pump speed, pump outlet pressure, and inlet pressure at that operating point are collected to calculate the head. The pump shaft power is directly measured using a torque meter or power analyzer. After recording the pump speed, corresponding head, and corresponding shaft power at each speed operating point, a speed-shaft power mapping function is constructed with the pump speed as the independent variable and the shaft power at that speed as the dependent variable. Specifically, the at least five speed operating points are sorted from smallest to largest speed value, forming several speed sub-intervals composed of adjacent speed points. For each speed sub-interval, a linear interpolation function is established with the speed at both ends of the interval as the independent variable and the corresponding shaft power as the dependent variable. That is, for the speed sub-interval [n]... i n i+1 Its linear interpolation function is P(n) = P i +(nn i )×(P i+1 -P i ) / (n i+1 -n i ), and set the endpoint speed values ​​n of each speed sub-range. i With n i+1 The corresponding endpoint shaft power value P i With P i+1and the interpolation slope factor (P) determined by the endpoint values. i+1 -P i ) / (n i+1 -n i The parameters, n, are stored in the adjustment module as a set of preset parameters in the form of a lookup table; where n is the pump speed currently collected by the sensing module. During real-time operation, the adjustment module determines the sub-range of pump speed based on the pump speed currently collected by the sensing module. It then retrieves the corresponding endpoint speed, endpoint shaft power, and interpolation slope factor from the lookup table and substitutes them into the linear interpolation function to calculate the pump shaft power under the current operating condition. This provides a basis for the accurate calculation of the current total power loss, thereby supporting the reliable determination of the subsequent total heat load and the reasonable allocation of the target cooling contribution of the dual cooling loops. In this embodiment, the linear interpolation function is used to calculate the corresponding pump shaft power based on the current pump speed. Where n... i and n i+1 These represent the speed values ​​of two adjacent speed operating points pre-calibrated by bench testing, forming a speed sub-range; P i and P i+1 Then, respectively at rotational speed n i and n i+1 The pump shaft power value obtained from the calibration is P(n), which is the pump shaft power under the current operating condition calculated by interpolation. This formula achieves a fast and accurate estimation of the pump shaft power by performing linear interpolation between adjacent calibration points where the current speed n falls, providing a key intermediate variable for subsequent accurate calculation of the current total power loss.

[0075] A102. The stator copper loss component in the current total power loss is corrected based on the stator winding temperature and the temperature coefficient of the copper resistance of the winding, to obtain the copper loss correction amount. The stator copper loss component is calculated from the motor current and the room temperature resistance of the stator winding. In this embodiment, the room temperature resistance R0 of the stator winding is measured at a reference temperature T0 when the motor leaves the factory and is preset in the adjustment module. The adjustment module obtains the current stator winding temperature Tw in real time, and uses ΔT = Tw - T0 as the winding temperature rise. The room temperature resistance R0 of the stator winding is multiplied by the temperature correction factor (1 + α). ΔT) is used to obtain the stator winding hot resistance Rw = R0 at the current actual operating temperature. (1+α ΔT), where α is the temperature coefficient of the winding copper resistance, and ΔT is the difference between the motor stator winding temperature and the reference temperature T0; then, the actual stator copper loss value P is obtained by multiplying the square of the motor current I by the hot resistance Rw of the stator winding. cu_w =I² Rw; Simultaneously, multiplying the square of the motor current I by the room-temperature resistance R0 of the stator winding yields the stator copper loss component P under room-temperature conditions.cu_0 =I² R0; based on the current actual stator copper loss value P cu_w Stator copper loss component P under normal temperature conditions cu_0 The difference ΔP cu =P cu_w -P cu_0 This copper loss correction amount is ΔP. cu This reflects the increase in copper loss caused by the temperature rise ΔT of the winding, allowing the determination of the total heat load to be dynamically adjusted according to the current change in the stator winding temperature Tw of the motor, avoiding underestimation of the actual heat generation due to ignoring the temperature effect of copper resistance.

[0076] A103. Compare the front bearing temperature with the normal temperature reference value of the front bearing under the current speed and load conditions to obtain the front bearing temperature rise deviation. In this embodiment, the normal temperature reference value of the front bearing is pre-established and preset in the adjustment module through bench testing during the motor factory test phase. Specifically, the establishment method is as follows: under standard cooling conditions, the water pump is run in M×N calibration conditions consisting of M calibration speeds and N calibration loads, where the calibration speed is denoted as nj (j=1,2). The rated load is denoted as Lk (k=1,2,…,N). Under each rated operating condition (nj,Lk), after the front bearing temperature reaches a steady state, the steady-state temperature value Tb_f0(nj,Lk) of the front bearing under that condition is recorded. The speed nj, load Lk, and steady-state temperature value Tb_f0(nj,Lk) corresponding to each rated operating condition are stored as a two-dimensional mapping table. This two-dimensional mapping table uses the current pump speed n and motor load L as indexes and the previous bearing normal temperature reference value Tb_ f0(n,L) is the output, pre-stored in the storage unit of the adjustment module; during real-time operation, the adjustment module acquires the front bearing temperature Tb_f, pump speed n, and motor load L calculated from motor current I and motor voltage U currently collected by the sensing module; using the current pump speed n and current motor load L as indexes, the two-dimensional mapping table is queried, and when (n,L) coincides with a certain calibration condition (nj,Lk), the corresponding Tb_f0(nj,Lk) is directly extracted as the front bearing... Normal temperature reference value; when (n,L) falls within the grid formed by four adjacent calibration conditions, the normal temperature reference value Tb_f0(n,L) of the front bearing, which matches the current condition, is calculated using bilinear interpolation based on the steady-state temperature values ​​of the front bearing corresponding to the four calibration conditions; the current front bearing temperature Tb_f is subtracted from the normal temperature reference value Tb_f0(n,L), and the resulting difference ΔTb_f=Tb_f-Tb_f0(n,L) is the front bearing temperature rise deviation.

[0077] A104. Similarly, the rear bearing temperature is compared with the normal temperature reference value of the rear bearing under the current speed and load conditions to obtain the rear bearing temperature rise deviation. In this embodiment, the normal temperature reference value of the rear bearing is synchronously recorded under each calibration condition (nj, Lk) after the rear bearing temperature reaches a steady state during the aforementioned test on the same bench, resulting in the rear bearing steady-state temperature value Tb_r0(nj, Lk). This forms a second two-dimensional mapping table with the current pump speed n and motor load L as indexes and the rear bearing normal temperature reference value Tb_r0(n, L) as the output, and is also pre-stored in the storage unit of the adjustment module. During real-time operation, the adjustment module obtains the rear bearing temperature Tb_r, pump speed n, and motor current I currently collected by the sensing module. The motor load L, calculated with the motor voltage U, is used as an index to query the second two-dimensional mapping table. When (n,L) coincides with a certain calibration condition, the corresponding Tb_r0(nj,Lk) is directly extracted as the reference value for the normal temperature of the rear bearing. When (n,L) falls within the grid of an adjacent calibration condition, the reference value for the normal temperature of the rear bearing, Tb_r0(n,L), is calculated using bilinear interpolation based on the steady-state temperature value of the rear bearing corresponding to each vertex of the grid. The difference ΔTb_r = Tb_r - Tb_r0(n,L) is the rear bearing temperature rise deviation. The rear bearing temperature rise deviation ΔTb_r only reflects the additional heat generated by the rear bearing due to abnormal factors such as poor lubrication, wear, or abnormal fit clearance. It is input together with the front bearing temperature rise deviation ΔTb_f into the pre-established bearing friction loss increment mapping relationship, providing a reliable input for the accurate calculation of bearing friction loss increment.

[0078] A105. Input the temperature rise deviation of the front bearing and the temperature rise deviation of the rear bearing into the pre-established bearing friction loss increment mapping relationship to obtain the bearing friction loss increment.

[0079] It should be noted that the bearing friction loss increment mapping relationship is used to map the front bearing temperature rise deviation ΔTb_f and the rear bearing temperature rise deviation ΔTb_r to the additional friction loss power generated by abnormal bearing heating, i.e., the bearing friction loss increment. In this embodiment, this mapping relationship is pre-established and pre-set in the adjustment module through bench testing during the motor factory testing phase. The specific construction includes:

[0080] A1051. Under the calibrated operating condition (nj, Lk), once the bearing temperature reaches a steady state and the lubrication condition is confirmed to be normal and the bearing clearance meets the design tolerance, the total shaft power of the water pump is measured synchronously by a torque meter. The copper loss calculated by the motor current I and the stator winding hot resistance Rw, as well as the pre-calibrated rotor iron loss and stray loss, are deducted from the power to obtain the sum of the front-to-rear bearing friction losses under normal conditions, Pbf0(nj, Lk), which is used as the reference friction loss for this operating condition.

[0081] A1052. Different degrees of poor lubrication conditions were artificially introduced at the front and rear bearings, including reducing the amount of grease injected and adding trace impurity particles to simulate wear. The operation was repeated until steady state under each calibrated condition (nj,Lk). The front bearing temperature Tb_f′, the rear bearing temperature Tb_r′, and the corresponding total friction loss Pbf′(nj,Lk) were recorded at this time. With the reference friction loss Pbf0(nj,Lk) as a reference, the bearing friction loss increment ΔPbf(nj,Lk) caused by abnormal temperature rise under the corresponding condition was obtained as Pbf′(nj,Lk) - Pbf0(nj,Lk).

[0082] A1053. Since the front and rear bearings share the same lubrication circuit and heat transfer is coupled, ΔPbf(nj,Lk) needs to be separated: Under the test condition of introducing an anomaly only to the front bearing while keeping the rear bearing normal, the temperature rise deviation ΔTb_f and the corresponding loss increment ΔPf(ΔTb_f) of the front bearing when the front bearing is abnormal alone are measured; Under the test condition of introducing an anomaly only to the rear bearing while keeping the front bearing normal, the temperature rise deviation ΔTb_r and the corresponding loss increment ΔPr(ΔTb_r) of the rear bearing when the rear bearing is abnormal alone are measured; ΔPf(ΔTb_f) and ΔPr(ΔTb_r) are used as independent friction loss increment mapping data for the front and rear bearings, respectively.

[0083] A1054. Perform linear regression or piecewise linear fitting on the front bearing temperature rise deviation ΔTb_f obtained in step 3 and the corresponding front bearing friction loss increment ΔPf to form the front bearing friction loss increment mapping function ΔPf=ff(ΔTb_f).

[0084] Similarly, the rear bearing temperature rise deviation ΔTb_r is linearly regressed or piecewise linearly fitted with the corresponding rear bearing friction loss increment ΔPr to form the rear bearing friction loss increment mapping function ΔPr=fr(ΔTb_r); these two mapping functions are stored in the adjustment module in the form of a lookup table or piecewise function.

[0085] A1055. During real-time operation, the adjustment module inputs the front bearing temperature rise deviation ΔTb_f calculated above into the mapping function ff to obtain the front bearing friction loss increment ΔPf; and inputs the rear bearing temperature rise deviation ΔTb_r into the mapping function fr to obtain the rear bearing friction loss increment ΔPr; the sum of the two, ΔPbf = ΔPf + ΔPr, is used as the bearing friction loss increment for subsequent superposition calculation of total heat load.

[0086] A106. The sum of the copper loss correction and the bearing friction loss increment is used as the loss correction, which is then added to the current total loss power to obtain the total heat load.

[0087] This embodiment effectively solves the problem of time lag and deviation in total heat load estimation caused by the spatial dispersion and mutual coupling of heat sources such as stator copper loss, bearing friction loss and rotor iron loss inside the motor, which are coupled together, resulting in the inability of the temperature at each temperature measurement point to directly map the real-time contribution of a single heat source, especially the lag in the response of bearing temperature to load changes. Specifically, the total power loss is calculated from the difference between the motor input power and the pump shaft power, allowing for the initial separation of relatively stable loss components such as iron loss and stray loss from copper loss and bearing friction loss that vary with operating conditions. Based on this, the stator copper loss component is corrected for temperature by introducing the winding copper resistance temperature coefficient, converting the nominal copper loss under normal temperature conditions into the actual copper loss value at the current operating temperature, thus eliminating the underestimation of heat generation due to increased copper resistance caused by winding temperature rise. Simultaneously, by comparing the front and rear bearing temperatures with normal temperature reference values ​​under the same speed and load conditions, the normal influence of operating condition changes on bearing temperature is eliminated, extracting only those reflecting poor lubrication, wear, or abnormal fit clearance. The temperature rise deviation is then converted into corresponding additional friction loss power through a pre-established bearing friction loss increment mapping relationship, realizing the independent capture and quantification of abnormal bearing heating. Finally, the sum of the copper loss correction and the bearing friction loss increment is added to the current total loss power as the loss correction, so that the total heat load can be dynamically adjusted according to the changes in winding temperature and bearing operating status. This ensures the accurate differentiation and tracking of the real-time contribution of each heat source, thereby providing a precise heat load input for the reasonable allocation of the target cooling contribution of the dual cooling circuit. It avoids insufficient or excessive cooling caused by heat load estimation deviation, and improves the energy efficiency level of the coordinated adjustment of the dual cooling system and the reliability of motor operation.

[0088] It should be noted that directly calculating the current cooling contribution of the internal cooling circuit using fixed cooling medium properties and collected inlet / outlet temperature differences and flow rates will encounter three levels of deviation: First, the specific heat capacity and density of the cooling medium change with temperature and medium composition, introducing property deviations when using fixed parameters; second, scaling or impurity deposition on the inner wall of the internal cooling circuit increases thermal resistance, resulting in less actual heat removal under the same flow rate and temperature difference than the theoretical value, and this change in thermal resistance cannot be directly perceived by the inlet / outlet temperature difference and flow rate; third, the accuracy of the flow sensor decreases under low flow rate conditions, and directly accepting the raw flow rate value will further exacerbate the calculation error. To solve the above problems, this embodiment provides a method for calculating the current cooling contribution of the internal cooling circuit, specifically:

[0089] B101. The inlet temperature, outlet temperature and flow rate of the internal cooling circuit are obtained by the sensing module.

[0090] B102. The average value of the inlet temperature and the outlet temperature is taken as the qualitative temperature of the cooling medium. Based on this qualitative temperature, the specific heat capacity and density of the cooling medium at the current temperature are obtained from the cooling medium property parameter mapping table pre-installed in the adjustment module by table lookup or interpolation. In this embodiment, the cooling medium property parameter mapping table is constructed as follows: for the type of medium actually transported by the water pump, under laboratory conditions, with water as the reference medium, a differential scanning calorimeter is used to set a sampling point every five degrees Celsius to measure the specific heat capacity at each sampling point temperature, and a vibrating tube density is used. The thermometer measures the density at each sampling point temperature. The temperature, specific heat capacity, and density values ​​corresponding to each sampling point are sequentially entered in the form of triplets to form a cooling medium property parameter mapping table, which is pre-stored in the read-only storage unit of the adjustment module. When the qualitative temperature is exactly equal to the temperature value of a certain sampling point, the specific heat capacity and density corresponding to that sampling point are directly extracted as output. When the qualitative temperature is between the temperature values ​​of two adjacent sampling points, the specific heat capacity and density corresponding to the qualitative temperature are calculated by linear interpolation based on the specific heat capacity and density values ​​corresponding to these two sampling points.

[0091] B103. Simultaneously, it is determined whether the flow rate of the internal cooling circuit is lower than a preset low flow rate threshold. If it is lower than the low flow rate threshold, a preset low flow rate correction coefficient is used to correct the flow rate, and the corrected flow rate is used as the flow rate value for calculation. Otherwise, the collected flow rate value is used directly. In this embodiment, the low flow rate threshold is set to 10% of the upper limit of the electromagnetic flowmeter's range, based on the inherent metering characteristics of the electromagnetic flowmeter and the factory bench calibration results. Due to the uneven flow velocity distribution and weak sensing signal of the electromagnetic flowmeter under low flow conditions, its measurement accuracy decreases significantly and nonlinear error increases. Through bench testing, a high-precision standard turbine flowmeter was used for comparison at different low flow rate points. It was found that in the range below 10% of the upper limit of the range, the measurement error began to exceed the system's allowable range. Therefore, this inflection point was set as the low flow rate threshold. When the real-time flow rate falls below this threshold, the system automatically calls a preset low flow rate correction coefficient for compensation. This ensures that the flow rate value used in the cooling contribution calculation accurately reflects the working state of the internal cooling circuit, thus providing accurate data support for subsequent heat load calculations and energy consumption optimization of the dual cooling circuits. The low flow rate correction coefficient is obtained as follows: During bench testing, a standard turbine flow meter with an accuracy class of not less than 0.2 is connected in series in the internal cooling circuit as a reference. The readings of the reference flow meter are recorded simultaneously at four low flow rate operating points: 3%, 5%, 7%, and 9% of the upper limit of the flow meter's range. The readings of the electromagnetic flowmeter are used to calculate the ratio of the reference flowmeter reading to the electromagnetic flowmeter reading at each operating point, which is used as the low flow correction coefficient for that operating point. A four-row, two-column correction coefficient table is formed with the electromagnetic flowmeter reading as the index and the low flow correction coefficient as the output, and is preset in the adjustment module. When it is determined in real time that the current electromagnetic flowmeter reading is lower than the low flow threshold, the two closest flow index points are found in the correction coefficient table with the current electromagnetic flowmeter reading as the index. The corresponding low flow correction coefficient is obtained by linear interpolation, and the current electromagnetic flowmeter reading is multiplied by the coefficient to obtain the corrected flow value.

[0092] B104. The theoretical cooling contribution of the internal cooling circuit is calculated as the product of the flow rate, density, specific heat capacity, and the difference between the outlet and inlet temperatures. In this embodiment, the difference between the outlet and inlet temperatures is obtained by subtracting the inlet temperature value collected by the first platinum resistance temperature sensor from the outlet temperature value collected by the second platinum resistance temperature sensor. If the difference is less than or equal to zero, it indicates that the internal cooling circuit has not removed heat from the motor side at that instant or the heat transfer direction is abnormal. At this time, the theoretical cooling contribution is set to zero, and an internal cooling circuit abnormality record containing the current timestamp, inlet temperature value, outlet temperature value, and flow rate value is generated. At the same time, an internal cooling circuit abnormality warning signal is triggered and pushed to the cloud platform to prompt the operator to check whether the cooling medium circulation is interrupted, whether there is a short circuit or bypass in the pipeline, or whether the temperature sensor is faulty. In this embodiment, the first platinum resistance temperature sensor and the second platinum resistance temperature sensor are both components of the sensing module and are respectively deployed on the internal cooling circuit of the fully sealed dual-cooling waterproof pump. The first platinum resistance thermometer (RTD) is installed at the inlet wall of the internal cooling circuit to collect the inlet temperature of the internal cooling circuit in real time. The second PRD is installed at the outlet wall of the internal cooling circuit to collect the outlet temperature of the internal cooling circuit in real time. Both use platinum resistance thermometers as temperature sensing elements. Utilizing the linear change in resistance with temperature, the temperature signal is converted into an electrical signal and transmitted to the control module. When calculating the current cooling contribution of the internal cooling circuit, the control module subtracts the inlet temperature value collected by the second PRD from the inlet temperature value collected by the first PRD. This difference between the outlet and inlet temperatures is used as the key temperature difference parameter for heat removal by the cooling medium in the internal cooling circuit.

[0093] B105. The estimated heat flow rate transferred from the motor to the cooling medium is obtained by calculating the differences between the motor stator winding temperature, front bearing temperature, and rear bearing temperature and the inlet temperature, multiplying each difference by its respective preset equivalent heat transfer coefficient, and then summing them. In this embodiment, the equivalent heat transfer coefficients Kcu (corresponding to the motor stator winding temperature), Kbf (corresponding to the front bearing temperature), and Kbr (corresponding to the rear bearing temperature) are all obtained during the motor's factory testing phase through a step-by-step isolation calibration method and preset in the adjustment module. The equivalent heat transfer coefficient characterizes the overall heat transfer capacity from each heat source to the cooling medium inside the shaft, through all heat transfer links such as insulation layer heat conduction, air gap or gap thermal resistance, metal wall heat conduction, and convection heat transfer, and is expressed in watts per degree Celsius (W / ℃).

[0094] The calibration process is as follows:

[0095] The first step is to calibrate Kcu. With the internal cooling circuit flow rate and inlet temperature maintained constant, the motor is operated at rated speed with increasing stator current at five operating points: 20%, 40%, 60%, 80%, and 100% of rated load. At each operating point, after the outlet temperature collected by the second platinum resistance thermometer reaches a steady state, the steady-state outlet temperature value is recorded. The product of the square of the effective value of the motor current I and the room-temperature resistance R0 of the stator winding is taken as the known stator copper loss Pcu_i for each operating point. Simultaneously, the total heat carried away by the cooling medium at each operating point is calculated as Qcool_i = Q × ρ × cp × ΔT, using the product of the internal cooling circuit flow rate Q, the specific heat capacity cp and density ρ of the cooling medium obtained from a qualitative temperature lookup table, and the difference between the outlet and inlet temperatures ΔT. Subtracting the pre-measured rotor iron loss Pfe and stray loss Pstray from the total heat Qcool_i yields the cooling heat contributed by stator copper loss at each operating point, Qcu_i = Qcool_i - Pfe - Pstray. Using the known stator copper loss Pcu_i at the five operating points as the independent variable and the corresponding cooling heat Qcu_i as the dependent variable, a univariate linear regression through the origin is performed using the least squares method. The slope of the regression line is Kcu, which physically represents the increase in heat absorbed by the cooling medium per unit increase in stator copper loss.

[0096] Taking a 110kW rated power motor as an example, under the conditions of a constant inlet temperature of 55℃ and a constant flow rate of 3.5m³ / h, the data at five load points are as follows: at 20% load, Pcu = 1.8kW, Qcu = 1.6kW; at 40% load, Pcu = 3.6kW, Qcu = 3.2kW; at 60% load, Pcu = 5.4kW, Qcu = 4.9kW; at 80% load, Pcu = 7.2kW, Qcu = 6.5kW; and at 100% load, Pcu = 9.0kW, Qcu = 8.1kW. Linear regression yields Kcu≈0.90W / ℃.

[0097] The second step is to calibrate Kbf and Kbr. With the stator current at zero (i.e., the motor is not energized), the rotor is driven to its rated speed using an external drive device. Three friction states—normal lubrication, semi-dry friction, and boundary friction—are simulated at the front and rear bearings by varying the amount of grease injected or adjusting the bearing end cap clamping force. Under each friction state, the temperatures of the front bearing (Tb_f), rear bearing (Tb_r), and outlet temperature are recorded synchronously after reaching a steady state. The total heat carried away by the cooling medium is calculated using the flow rate, specific heat capacity, density, and outlet temperature rise of the internal cooling circuit. Since the stator current is zero, the total heat carried away by the cooling medium at this point is the sum of the friction losses of the front and rear bearings. By conducting three sets of experiments that changed only the friction state of the front bearing while keeping the state of the rear bearing constant, a linear regression relationship was established between the front bearing temperature Tb_f and the total heat carried away by the cooling medium, and the regression slope is Kbf. Similarly, by conducting three sets of experiments that changed only the friction state of the rear bearing while keeping the state of the front bearing constant, a linear regression relationship was established between the rear bearing temperature Tb_r and the total heat carried away by the cooling medium, and the regression slope is Kbr.

[0098] The calibrated Kcu, Kbf, and Kbr are preset as constant coefficients in the storage unit of the adjustment module. During real-time operation, the adjustment module multiplies Kcu by the difference between the motor stator winding temperature and the inlet temperature, Kbf by the difference between the previous bearing temperature and the inlet temperature, and Kbr by the difference between the subsequent bearing temperature and the inlet temperature. The sum of these three results yields the estimated heat flow rate transferred by the motor's lateral cooling medium, which is used to calculate the thermal resistance correction coefficient of the subsequent internal shaft cooling circuit.

[0099] B106. The ratio of the estimated heat flow to the theoretical cooling contribution is used as the thermal resistance correction coefficient under the current operating conditions. When the ratio is greater than 1, it indicates that the thermal resistance has increased due to scaling or impurity deposition on the inner wall of the shaft cooling circuit, and the actual cooling capacity is lower than the theoretical value. In this embodiment, the adjustment module continuously calculates the thermal resistance correction coefficient at a preset sampling period. If the thermal resistance correction coefficient is greater than 1 for 60 consecutive sampling periods, and the slope obtained by linear regression of the thermal resistance correction coefficient sequence within the 60 sampling periods is greater than 0, that is, it shows a statistically significant monotonically increasing trend, then it is further determined that there is progressive blockage or increased scaling in the shaft cooling circuit, triggering a maintenance warning signal for the shaft cooling circuit. The maximum value, average value, and current value of the thermal resistance correction coefficient within the 60 sampling periods are packaged into a warning information package and pushed to the cloud platform to prompt the operator to arrange cleaning or maintenance. In this embodiment, the preset sampling period is determined as follows: During the system factory commissioning phase, a step-change cooling flow rate is first applied to the internal cooling circuit, and the time required for the outlet temperature collected by the second platinum resistance temperature sensor to reach 63.2% of the steady-state value from the start of the step change is recorded as the thermal response time constant τ of the internal cooling circuit; simultaneously, the maximum data refresh periods of the first platinum resistance temperature sensor, the second platinum resistance temperature sensor, and the electromagnetic flowmeter are obtained respectively, and the maximum value among the three is taken as the maximum refresh period T. max The product of the thermal response time constant τ and the preset empirical proportionality coefficient K is multiplied by the maximum refresh period T. max The two values ​​are compared, and the larger value is taken as the preset sampling period. This sampling period is then preset in the storage unit of the adjustment module as a configuration parameter.

[0100] B107. Divide the theoretical cooling contribution by the thermal resistance correction coefficient to obtain the corrected current cooling contribution of the in-shaft cooling circuit. In this embodiment, if the thermal resistance correction coefficient is equal to 1 or within the tolerance range of 1 ± 0.05, it is determined that the thermal resistance has not changed significantly, and the theoretical cooling contribution is directly used as the current cooling contribution of the in-shaft cooling circuit. If the thermal resistance correction coefficient is less than 0.95, it indicates that the estimated heat flow is greater than the theoretical cooling contribution. At this time, the abnormality is recorded as a system deviation event and stored in the local log. The theoretical cooling contribution is still used as the current cooling contribution of the in-shaft cooling circuit. At the same time, a self-test prompt signal is triggered and pushed to the cloud platform. The self-test prompt signal includes the current thermal resistance correction coefficient value, the theoretical cooling contribution value, the estimated heat flow value, and the preset values ​​of the three equivalent heat transfer coefficients, prompting the operator to calibrate and check the calibration validity of the equivalent heat transfer coefficient or the measurement drift of the temperature sensor.

[0101] This embodiment introduces a mechanism for dynamically looking up cooling medium properties with varying qualitative temperatures. It uses the average inlet and outlet temperatures to index a pre-set cooling medium property parameter mapping table in real time to obtain the true specific heat capacity and density under current operating conditions. This eliminates the property deviations introduced by fixed parameters in existing technologies when the medium temperature or composition fluctuates. Simultaneously, by setting a low flow rate threshold and applying a low flow rate correction coefficient based on bench test calibration to flow rates below the threshold, it compensates for the underestimation of flow rate caused by decreased flow sensor measurement accuracy under low flow rate conditions, making the calculated flow rate values ​​closer to the true values. Furthermore, by... The ratio of the theoretical cooling contribution to the estimated heat flow calculated by inverting the temperature difference between the heat sources on the motor side and the inlet temperature is used to obtain the real-time thermal resistance correction coefficient. The theoretical cooling contribution is then divided by this coefficient to output the corrected current cooling contribution. This allows the increase in thermal resistance caused by scaling or impurity deposition on the inner wall of the shaft cooling circuit to be dynamically sensed and incorporated into the correction calculation of the cooling contribution. This accurately reflects the actual cooling capacity of the shaft cooling circuit, providing accurate input for the reasonable allocation of the target cooling contribution of the dual cooling circuits. It avoids target allocation deviations caused by overestimating the cooling capacity of the shaft cooling circuit and the resulting risk of motor overheating.

[0102] It should be noted that directly calculating the current cooling contribution of the external forced air cooling circuit based on the difference between the ambient temperature at the far end of the air inlet and the single-point temperature of the heat sink fins, as well as the rotational speed of the cooling fan, will encounter three levels of deviation: First, the rotational speed of the cooling fan only reflects the speed of the fan itself, while the actual effective cooling airflow through the heat sink fins is also significantly affected by the airflow resistance, the degree of dust accumulation on the fin surface, and the blockage of the air inlet filter. When there is dust accumulation on the fin surface or the filter is partially blocked, the actual airflow at the same rotational speed has been greatly reduced. Calculating the airflow solely based on the rotational speed will be highly inaccurate. First, the actual heat dissipation capacity of external air cooling is overestimated. Second, the surface temperature of the heat sink fins is gradient-distributed along the airflow direction, and the temperature at a single point cannot represent the average temperature of the entire fin heat exchange surface. Calculating heat dissipation based on the difference between a single point temperature and the ambient temperature introduces errors due to temperature field inhomogeneity. Third, the ambient temperature near the air inlet may experience localized temperature rise due to the motor's own heat dissipation or the heat plume from nearby heat sources. The collected ambient temperature does not truly represent the ambient temperature at the far end of the cooling air inlet, leading to an underestimation of the calculated temperature difference and the cooling contribution. To resolve these issues, please refer to [link to relevant documentation]. Figure 2 This embodiment provides a method for calculating the current cooling contribution of an external forced air cooling circuit, specifically as follows:

[0103] B201. Obtain the cooling fan speed. Based on the mapping relationship between fan speed and theoretical air volume preset in the adjustment module, convert the current cooling fan speed into theoretical cooling air volume. In this embodiment, the mapping relationship between fan speed and theoretical air volume is constructed as follows: During the factory test phase, under the conditions that the air inlet filter is removed and the heat dissipation fins are in a factory clean state, nine speed calibration points are selected at 10% intervals within the range of 20% to 100% of the rated speed of the cooling fan. At each speed calibration point, after the speed stabilizes, the windward surface of the heat dissipation fins is divided into nine rectangular sub-regions of equal area. A thermal anemometer is used to measure the local wind speed at the geometric center point of each sub-region. The arithmetic mean of the nine local wind speeds is multiplied by the total area of ​​the windward surface to obtain the theoretical cooling air volume corresponding to the speed calibration point. The nine speed calibration points and the corresponding theoretical cooling air volume are stored as a lookup table and preset in the adjustment module. During real-time operation, the adjustment module obtains the current cooling fan speed. If the current speed is equal to the speed at a certain calibration point, the corresponding theoretical cooling air volume is directly extracted. If the current speed is between two adjacent calibration point speeds, the corresponding theoretical cooling air volume is calculated by linear interpolation.

[0104] B202. Obtain the pressure difference value across the air inlet filter. Compare the pressure difference value with the reference pressure difference value pre-set in the adjustment module under the condition of a clean filter to obtain the filter clogging pressure difference increment. In this embodiment, a pressure tap of a first micro differential pressure sensor is installed 5 cm upstream of the air inlet filter from the windward side of the filter, and a pressure tap of a second micro differential pressure sensor is installed 5 cm downstream of the air inlet filter from the leeward side of the filter. The upper limit of the range of both micro differential pressure sensors is selected as 1.5 times the theoretical maximum pressure drop when the filter is completely clogged. The reference pressure difference value is the difference between the reading of the first micro differential pressure sensor and the reading of the second micro differential pressure sensor under the conditions of a clean filter at the factory, the cooling fan running at its rated speed, and no external obstruction at the air inlet. This reference pressure difference value is pre-set in the adjustment module. During real-time operation, the adjustment module obtains the current difference between the readings of the first and second micro differential pressure sensors as the current pressure difference value. Subtracting the reference pressure difference value from the current pressure difference value yields the filter clogging pressure difference increment.

[0105] B203. Simultaneously, the fin temperature values ​​collected from multiple preset temperature measurement points on the surface of the heat dissipation fins are obtained, and the arithmetic mean of the multiple fin temperature values ​​is taken as the average fin temperature. In this embodiment, the heat dissipation fins are divided into three sections along the airflow direction: an inlet section, a middle section, and an outlet section. The inlet section is located 5 mm inward from the windward edge of the fin, the outlet section is located 5 mm inward from the air outlet edge of the fin, and the middle section is located at the geometric center of the inlet and outlet sections. Three temperature measurement points (upper, middle, and lower) are evenly arranged along the height direction of the fin on each section. The upper temperature measurement point is 3 mm from the top edge of the fin, the lower temperature measurement point is 3 mm from the root of the fin, and the middle temperature measurement point is located at the center of the upper and lower temperature measurement points. There are a total of 9 temperature measurement points. The temperature of each temperature measurement point is collected by a patch-type platinum resistance temperature sensor. The space between the sensor's sensitive element and the fin surface is filled with thermally conductive silicone grease and covered with a heat insulation pad. The sum of the temperature values ​​collected from the 9 temperature measurement points is divided by 9, and the result is the average fin temperature.

[0106] B204. Based on the filter clogging pressure difference increment, obtain the airflow attenuation coefficient under the current operating condition from the mapping relationship between the filter clogging pressure difference increment and the airflow attenuation coefficient preset in the adjustment module; multiply the theoretical cooling airflow by the airflow attenuation coefficient to obtain the effective cooling airflow. In this embodiment, the mapping relationship between the filter clogging pressure difference increment and the airflow attenuation coefficient is established during the factory test stage through the following steps: uniformly cover the filter surface with shielding materials covering 10%, 20%, 30%, 40%, and 50% of the total filter area to simulate five different degrees of clogging; under each clogging state, run the cooling fan at five speed points: 20%, 40%, 60%, 80%, and 100% of the rated speed range. At each speed point, divide the windward surface of the heat dissipation fins into nine rectangular sub-regions of equal area. Use a thermal anemometer to measure the local wind speed at the geometric center point of each sub-region, and take the arithmetic mean of the nine local wind speeds as... The average wind speed on the windward side is multiplied by the total area of ​​the windward side to obtain the current actual air volume. The ratio of the actual air volume at the blockage state and speed point to the theoretical air volume at the same speed point in the clean state is calculated and used as the air volume attenuation coefficient for that state. The filter blockage pressure difference increment corresponding to each blockage state and the arithmetic mean of the air volume attenuation coefficients corresponding to all speed points measured under each blockage state are stored in a mapping table and preset in the adjustment module. During real-time operation, the current filter blockage pressure difference increment is looked up in the table. If it is equal to the pressure difference increment of a preset blockage state, the corresponding air volume attenuation coefficient is directly extracted. If it is between the pressure difference increments of two preset blockage states, the air volume attenuation coefficient is obtained by linear interpolation.

[0107] B205. Obtain the ambient temperature at the far end of the air inlet. The ambient temperature at the far end of the air inlet is collected by an independent ambient temperature sensor installed away from the heat-affected zone of the motor. In this embodiment, the independent ambient temperature sensor is installed upstream of the water pump air inlet, at a vertical distance of 3 times the equivalent diameter of the air inlet from the windward side of the air inlet filter. The sensor uses a platinum resistance thermometer element and is externally equipped with a radiation shield composed of two concentric cylinders. An air gap layer is formed between the inner and outer cylinders. The outer surface of the outer cylinder is coated with a metallic reflective coating with a reflectivity of not less than 0.85 to prevent direct sunlight and infrared heat radiation from nearby equipment from directly heating the temperature sensing element. The independent ambient temperature sensor transmits the collected signal to the adjustment module through a shielded cable.

[0108] B206. The difference between the average fin temperature and the ambient temperature at the far end of the air inlet is taken as the effective heat exchange temperature difference. In this embodiment, if the difference between the average fin temperature and the ambient temperature at the far end of the air inlet is less than or equal to 0, it is determined that the external air cooling failed to remove heat from the heat dissipation fins at that instant. The current cooling contribution of the external forced air cooling circuit is directly set to 0. At the same time, an abnormal status record is generated, including the current timestamp, the average fin temperature value, the ambient temperature at the far end of the air inlet value, the cooling fan speed value, and the filter blockage pressure difference increment value, triggering an abnormality warning signal for the external air cooling circuit.

[0109] B207. Based on the effective cooling airflow, the isobaric specific heat capacity of air, the air density, and the effective heat exchange temperature difference, calculate the current cooling contribution of the external forced air cooling circuit. In this embodiment, the two physical property parameters, the isobaric specific heat capacity and the air density, are dynamically obtained by querying a physical property parameter correspondence table pre-installed in the adjustment module and supplementing it with linear interpolation, based on the real-time ambient temperature at the air inlet collected by the sensing module. Specifically, during the system factory commissioning phase, using dry air at standard atmospheric pressure as the medium, a sampling temperature point is set every 5°C within a temperature range of -10°C to 60°C. The isobaric specific heat capacity and density values ​​corresponding to each sampling temperature point are obtained by consulting standard physical property data manuals or by actual measurement. A isobaric specific heat capacity-temperature correspondence table is constructed for each sampling temperature point and its corresponding isobaric specific heat capacity value, and a density-temperature correspondence table is constructed for each sampling temperature point and its corresponding density value. Both tables are indexed by temperature values ​​and output with the corresponding physical property parameter values, and are pre-installed in the read-only storage unit of the adjustment module. During real-time operation, the adjustment module uses the current ambient temperature T_amb_remote at the far end of the air inlet, collected by the sensing module, as a query index to match it in the constant pressure specific heat capacity-temperature correspondence table: if T_amb_remote matches a certain sampling temperature point T in the table... i If they are equal, then directly extract the isobaric specific heat capacity value cp(T) corresponding to that temperature point. i ) as cp_air; if T_amb_remote is between two adjacent sampling temperature points Ti and T i+1 Between, and T i < T_amb_remote < T i+1 Then, using the linear interpolation formula cp_air = cp(T) i ) + [cp(T i+1 -cp(T) i )] × (T_amb_remote - T i ) / (T i+1 - T i The specific heat capacity of air at constant pressure under the current operating conditions is calculated. Similarly, the air density ρ_air under the current operating conditions is obtained from the density-temperature correspondence table using the same matching and interpolation logic. The dynamically obtained cp_air and ρ_air are substituted into the calculation formula for the current cooling contribution of the external forced air cooling circuit, that is, multiplied by the effective cooling air volume and the effective heat exchange temperature difference, to ensure the accuracy of the heat dissipation power estimation under different ambient temperatures.

[0110] This embodiment introduces a filter clogging pressure difference detection and airflow attenuation coefficient correction mechanism. The difference between the real-time pressure difference across the filter and the baseline pressure difference in a clean state is used as the clogging pressure difference increment. The airflow attenuation coefficient is dynamically obtained from a preset mapping relationship between the clogging pressure difference increment and the airflow attenuation coefficient. This corrects the theoretical cooling airflow obtained from the fan speed conversion to obtain the effective cooling airflow, eliminating the overestimation of actual heat dissipation capacity caused by neglecting filter clogging and fin dust accumulation in existing technologies. Simultaneously, multiple measuring... The arithmetic mean of the measured temperatures is used as the average fin temperature, replacing the single-point temperature method used in existing technologies. This eliminates the error introduced by using a single-point temperature to represent the average temperature of the entire heat exchange surface due to the gradient distribution of the fin surface temperature field along the airflow direction. Furthermore, by moving the ambient temperature sampling point from the near end of the air inlet to a location far upstream of the motor's heat-affected zone and installing a radiation shield, the ambient temperature at the far end of the air inlet is used instead of the near-end ambient temperature affected by the motor's own heat dissipation and the heat plume from nearby heat sources. This eliminates the problem of underestimating the calculated temperature difference caused by localized temperature rises in the ambient temperature. The synergistic effect of these three measures ensures that the calculated current cooling contribution of the external forced air cooling circuit accurately reflects its actual heat dissipation capacity. This provides a reliable input for the reasonable allocation of the target cooling contribution of the dual cooling circuits, preventing the overestimation of the current cooling contribution of the external air cooling circuit, which could lead to an excessive cooling load on the internal cooling circuit and the risk of motor overheating.

[0111] Example 3

[0112] It should be noted that if the target cooling contribution of the in-shaft cooling circuit and the external forced air cooling circuit is determined by directly allocating the total heat load proportionally or by a fixed weight allocation, three problems will arise: First, the energy consumption function of the in-shaft cooling circuit and the external forced air cooling circuit and their output cooling contribution are not linearly related. Allocating according to a fixed ratio cannot achieve the lowest total cooling energy consumption, and may cause unnecessary energy waste under some operating conditions. Second, the cooling effects of the two circuits on different temperature measurement points are cross-coupled. That is, the change in the cooling flow rate of the in-shaft circuit not only affects the temperature of the motor stator winding, but also affects the temperature of the front and rear bearings. The cooling effect of the external air cooling on the heat dissipation fins indirectly changes the heat dissipation boundary of the motor housing and bearings. Using a fixed allocation rule cannot guarantee that the temperature rise of each temperature measurement point does not exceed the corresponding safety threshold. Third, when the total heat load exceeds the sum of the maximum cooling capacity of the two circuits, the fixed allocation strategy lacks a boundary handling mechanism for empty feasible domains, and will output an unexecutable adjustment command. To address the aforementioned issues, this embodiment provides a method for allocating the target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit, specifically as follows:

[0113] C101. Obtain the total heat load, the current cooling contribution of the in-shaft cooling circuit, and the current cooling contribution of the external forced air cooling circuit; within the preset in-shaft cooling flow rate adjustment range, discretize into multiple candidate in-shaft cooling contributions according to the first step length; within the preset cooling fan speed adjustment range, discretize into multiple candidate external air cooling contributions according to the second step length; combine the multiple candidate in-shaft cooling contributions and the multiple candidate external air cooling contributions in pairs to form a candidate allocation scheme set. In this embodiment, the lower limit of the in-shaft cooling flow rate adjustment range is 0.5 times the current cooling contribution of the in-shaft cooling circuit, and the upper limit is the maximum output cooling contribution of the in-shaft cooling circuit. The first step length is set to 5% of the difference between the upper and lower limits. The in-shaft cooling flow rate adjustment range is uniformly discretized into 21 candidate in-shaft cooling contributions according to this step length. The lower limit of the cooling fan speed adjustment range is 0.5 times the current cooling contribution of the external forced air cooling circuit, and the upper limit is the maximum output cooling contribution of the external forced air cooling circuit. The second step length is set to 5% of the difference between the upper and lower limits. The cooling fan speed adjustment range is uniformly discretized into 21 candidate external air cooling contributions according to this step length. The 21 candidate in-shaft cooling contributions and the 21 candidate external air cooling contributions are combined in pairs to form a candidate allocation scheme set of a total of 441 candidate allocation schemes.

[0114] C102. For each candidate allocation scheme in the candidate allocation scheme set, based on the temperature rise response matrix preset in the adjustment module, calculate the predicted temperature rise of the motor stator winding, the predicted temperature rise of the front bearing, and the predicted temperature rise of the rear bearing under that candidate allocation scheme. In this embodiment, the temperature rise response matrix is ​​a 3x2 coefficient matrix, denoted as A, where A11 represents the influence coefficient of the unit in-shaft cooling contribution increment on the temperature rise of the motor stator winding, A12 represents the influence coefficient of the unit external air cooling contribution increment on the temperature rise of the motor stator winding, A21 and A22 respectively represent the influence coefficients on the temperature rise of the front bearing, and A31 and A32 respectively represent the influence coefficients on the temperature rise of the rear bearing; the temperature rise response matrix is ​​calibrated during the factory test stage through the following steps: when the motor is running at rated speed at 50... Under the baseline operating condition of % rated load, the first step is to maintain the external air cooling contribution unchanged and increase the shaft cooling contribution from the current value in five steps, each step increasing by 10% of the maximum output cooling contribution. After each step, wait for the temperature at each measuring point to reach a steady state again, and record the temperature rise changes of the motor stator winding, the front bearing, and the rear bearing. Perform a univariate linear regression on the five temperature rise changes with the increment of the shaft cooling contribution, and the regression slopes are A11, A21, and A31. The second step is to maintain the shaft cooling contribution... Without changing the current value, the external air cooling contribution is increased in five steps, each time by 10% of the maximum output cooling contribution. After each step, the temperature at each measuring point is allowed to reach a steady state again, and the temperature rise is recorded. A univariate linear regression is performed on the temperature rise change and the increase in the external air cooling contribution, and the regression slopes are A12, A22, and A32. The six coefficients obtained from the calibration are preset in the adjustment module. During real-time calculation, the internal cooling contribution is obtained by subtracting the current internal cooling contribution from the candidate internal cooling contribution, resulting in the internal increment ΔPc. The cooling contribution is subtracted from the current external air cooling contribution to obtain the air cooling side increment ΔPf_w. The predicted temperature rise is calculated as follows: the predicted temperature rise of the motor stator winding is equal to the current measured temperature rise of the motor stator winding plus the product of A11 and ΔPc plus the product of A12 and ΔPf_w; the predicted temperature rise of the front bearing is equal to the current measured temperature rise of the front bearing plus the product of A21 and ΔPc plus the product of A22 and ΔPf_w; the predicted temperature rise of the rear bearing is equal to the current measured temperature rise of the rear bearing plus the product of A31 and ΔPc plus the product of A32 and ΔPf_w.

[0115] C103. Eliminate any candidate allocation scheme in the candidate allocation scheme set whose predicted temperature rise exceeds its corresponding safety threshold, and retain candidate allocation schemes whose predicted temperature rise does not exceed the corresponding safety threshold, thus forming a feasible allocation scheme set. In this embodiment, the safety threshold corresponding to the motor stator winding temperature is set as the maximum allowable long-term operating temperature of the insulation class used in the motor minus a safety margin of 5°C. The setting is based on the following: the insulation class used in the motor (such as F or H) specifies its maximum allowable long-term operating temperature in relevant national standards. This temperature value is the thermal stability limit of the insulation material during continuous operation within its rated life. However, in engineering practice, to cope with uncertainties such as transient overload of the motor, hysteresis in the cooling system, and measurement deviations of temperature sensors, a certain thermal protection buffer zone needs to be reserved based on the maximum allowable temperature of the insulation class. The 5°C safety margin is determined based on industry-standard experience in motor thermal protection engineering and the design specifications of this pump motor manufacturer. Its value effectively prevents accelerated thermal aging and reduced insulation life due to prolonged exposure to near-limit temperatures, while avoiding excessive margin that could lead to premature cooling system intervention and unnecessary energy waste. This safety margin value is preset as a fixed parameter in the adjustment module during the system's factory commissioning phase, forming, together with the maximum allowable temperature for the corresponding insulation class, a safe boundary that the temperature rise of the motor stator windings cannot exceed. The safety thresholds corresponding to the front and rear bearing temperatures are respectively set as the maximum allowable operating temperature listed in the bearing manufacturer's specifications minus the 5°C safety margin; all three safety thresholds are preset as constant values ​​in the adjustment module.

[0116] C104. For each feasible allocation scheme in the set of feasible allocation schemes, based on the energy consumption function of the in-shaft cooling circuit and the energy consumption function of the external forced air cooling circuit preset in the adjustment module, calculate the estimated energy consumption of the in-shaft cooling circuit and the estimated energy consumption of the external forced air cooling circuit under the feasible allocation scheme, and take the sum of the two as the estimated total cooling energy consumption of the feasible allocation scheme. In this embodiment, the energy consumption function of the in-shaft cooling circuit is determined during the factory test phase by selecting 10 operating points at equal intervals within the range of the minimum output value to the maximum output value of the in-shaft cooling contribution. At each operating point, the in-shaft cooling flow rate is adjusted to make the in-shaft cooling contribution reach the set value. After the operating conditions stabilize, the input power of the in-shaft cooling circuit drive motor is measured by a power meter. The 10 sets of in-shaft cooling contribution and the corresponding input power data are fitted with a quadratic polynomial using the least squares method to obtain the energy consumption function of the in-shaft cooling circuit Ec=a2. Pc²+a1 Pc+a0, where Pc is the contribution of internal cooling, and a2, a1, and a0 are fitting coefficients; the energy consumption function of the external forced air cooling circuit is obtained by the same method, taking 10 operating points within the range of external air cooling contribution, measuring the cooling fan drive power, and fitting the external forced air cooling circuit energy consumption function Ef=b2. Pf²+b1 Pf+b0, where Pf is the contribution of external air cooling, and b2, b1, b0 are fitting coefficients; the two sets of fitting coefficients a2, a1, a0 and b2, b1, b0 are pre-set in the adjustment module; during real-time operation, the candidate in-shaft cooling contribution in the feasible allocation scheme is substituted into Ec to calculate the estimated energy consumption of the in-shaft cooling circuit, and the candidate external air cooling contribution is substituted into Ef to calculate the estimated energy consumption of the external forced air cooling circuit.

[0117] C105. From the set of feasible allocation schemes, select the feasible allocation scheme with the minimum estimated total cooling energy consumption as the optimal allocation scheme. The candidate in-shaft cooling contribution of this optimal allocation scheme is taken as the target cooling contribution of the in-shaft cooling circuit, and the candidate external air cooling contribution of this optimal allocation scheme is taken as the target cooling contribution of the external forced air cooling circuit. In this embodiment, if there are multiple feasible allocation schemes with equal and minimum estimated total cooling energy consumption, the scheme with the smaller estimated energy consumption of the in-shaft cooling circuit is preferentially selected as the optimal allocation scheme. This prioritizes utilizing the cooling capacity carried by the transport medium itself and reduces the mechanical wear of the external air-cooled fan blades under the same total energy consumption conditions. After determining the target cooling contribution, the target cooling contribution of the internal cooling circuit is used to calculate the corresponding target internal cooling flow rate based on the difference between the inlet and outlet temperatures of the internal cooling circuit and the specific heat capacity and density of the cooling medium obtained from B102, and then converted into an internal cooling flow rate adjustment command. The target cooling contribution of the external forced air cooling circuit is used to calculate the corresponding target cooling fan speed based on the mapping relationship between the fan speed and the theoretical air volume, the air volume attenuation coefficient obtained from B204, the air density and constant pressure specific heat capacity obtained from B207, and the effective heat exchange temperature difference determined from B206, and then converted into a cooling fan speed adjustment command. The internal cooling flow rate adjustment command and / or the cooling fan speed adjustment command are then executed.

[0118] C106. If the feasible allocation scheme set is empty after the candidate allocation scheme set is eliminated in C103, then the boundary processing strategy is executed: the maximum output cooling contribution of the in-shaft cooling circuit and the maximum output cooling contribution of the external forced air cooling circuit are respectively used as the target cooling contribution and converted into corresponding adjustment command outputs, while triggering a cooling capacity insufficient alarm signal. In this embodiment, the cooling capacity insufficient alarm signal includes the current total heat load value, the maximum output cooling contribution of the in-shaft cooling circuit, the maximum output cooling contribution of the external forced air cooling circuit, the difference between the sum of the maximum cooling capacity of the two circuits and the total heat load, and the current timestamp. The above information is encapsulated and pushed to the cloud platform to prompt the operator that the current total heat load has exceeded the sum of the maximum cooling capacity of the two cooling circuits, and that the water pump operating load needs to be reduced or the cooling circuit needs to be checked for serious blockage faults.

[0119] It should be noted that if, after obtaining the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit, they are directly converted into in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands respectively using a fixed linear ratio, two problems will arise: First, the relationship between the in-shaft cooling contribution and the in-shaft cooling flow rate is not a constant coefficient, but is also affected by the specific heat capacity, density, and inlet / outlet temperature difference of the current cooling medium. If the real-time changes of these factors are ignored, the converted in-shaft cooling flow rate adjustment command will not accurately achieve the target cooling contribution. Second, the relationship between the external air cooling contribution and the cooling fan speed is also affected in real-time by factors such as the effective heat exchange temperature difference, air density, air constant pressure specific heat capacity, and airflow attenuation coefficient. Directly converting them with a fixed ratio will cause the cooling fan speed adjustment command to deviate from the actual required value. To solve the above problems, this embodiment provides a specific implementation method for converting the target cooling contribution into adjustment commands and executing them, specifically as follows:

[0120] D101. Obtain the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit as determined by C105; obtain the specific heat capacity and density of the cooling medium under the current operating conditions by B102; obtain the current inlet and outlet temperatures of the in-shaft cooling circuit by the sensing module, and calculate the difference between the inlet and outlet temperatures; divide the target cooling contribution of the in-shaft cooling circuit by the product of the specific heat capacity, the density, and the difference between the inlet and outlet temperatures to obtain the target in-shaft cooling flow rate; match the target in-shaft cooling flow rate with the mapping relationship between the in-shaft cooling flow rate and the water pump speed preset in the adjustment module to obtain the corresponding target in-shaft cooling water pump speed, and convert the target in-shaft cooling water pump speed into an in-shaft cooling flow rate adjustment command output. In this embodiment, if the difference between the inlet and outlet temperatures is less than or equal to 0, the generation of the in-shaft cooling flow rate adjustment command is paused, the current in-shaft cooling flow rate remains unchanged, and an in-shaft cooling circuit abnormality warning signal is triggered.

[0121] D102. Obtain the effective heat exchange temperature difference under the current operating condition from B206, obtain the air density and air specific heat capacity under the current operating condition from B207, and obtain the airflow attenuation coefficient under the current operating condition from B204. Divide the target cooling contribution of the external forced air cooling circuit by the product of the air density, the air specific heat capacity under the current operating condition, and the effective heat exchange temperature difference to obtain the target effective airflow of the external air cooling circuit. Divide the target effective airflow of the external air cooling circuit by the airflow attenuation coefficient to obtain the target theoretical airflow of the external air cooling circuit. Perform a reverse matching between the target theoretical airflow of the external air cooling circuit and the mapping relationship between the fan speed and the theoretical airflow preset in the adjustment module to obtain the corresponding target speed of the cooling fan, and convert the target speed of the cooling fan into a cooling fan speed adjustment command output. In this embodiment, if the effective heat exchange temperature difference is less than or equal to 0, the generation of the cooling fan speed adjustment command is paused, the current cooling fan speed remains unchanged, and an abnormal warning signal for the external air cooling circuit is triggered.

[0122] D103. Execute the shaft cooling flow rate adjustment command and / or the cooling fan speed adjustment command. In this embodiment, when the adjustment module executes the adjustment command, it prioritizes the shaft cooling flow rate adjustment command. After the shaft cooling flow rate reaches the target flow rate and stabilizes for a preset time, it then executes the cooling fan speed adjustment command. This prioritizes utilizing the cooling capacity carried by the conveying medium itself for adjustment, supplemented by external air cooling, to avoid system oscillation caused by simultaneous large-scale adjustments of both circuits.

[0123] In this embodiment, when converting the target cooling contribution of the internal cooling circuit into an internal cooling flow rate adjustment command, the specific heat capacity, density, and inlet / outlet temperature difference of the cooling medium under the current operating conditions are obtained in real time. The target cooling contribution is then divided by the product of these three factors to calculate the target internal cooling flow rate, rather than using a fixed linear ratio for direct conversion. This eliminates conversion deviations caused by changes in the physical properties of the cooling medium with temperature and fluctuations in the inlet / outlet temperature difference, ensuring that the internal cooling flow rate adjustment command accurately achieves the target cooling contribution. When converting the target cooling contribution of the external forced air cooling circuit into a cooling fan speed adjustment command, the effective heat exchange temperature difference, air density, and air constant pressure specific heat capacity under the current operating conditions are obtained in real time. The system incorporates an airflow attenuation coefficient, and calculates the target effective airflow, target theoretical airflow, and target cooling fan speed step by step based on the target cooling contribution. This fully considers the real-time impact of airflow attenuation, changes in air properties, and fluctuations in heat exchange temperature difference on the conversion relationship, eliminating command deviations caused by fixed-proportion conversion. Simultaneously, during the command execution phase, a sequential control strategy is adopted that prioritizes the execution of the in-shaft cooling flow rate adjustment command and waits for it to stabilize before executing the cooling fan speed adjustment command. This strategy prioritizes the use of the cooling capacity carried by the conveying medium itself for adjustment, supplemented by external air cooling. This avoids system oscillations caused by simultaneous large-scale adjustments in both loops, achieving a smooth transition in the adjustment process and minimizing cooling energy consumption.

[0124] Example 4

[0125] It should be noted that directly using the instantaneous amplitude of the leakage signal at the sealing joint, the current reading of the internal humidity of the motor, and the cumulative running time as inputs, and judging sealing failure through a single threshold comparison or a simple linear model, will face three levels of lower-level technical problems: First, the signal amplitude output by the leakage sensor at the sealing joint during the micro-leakage stage is extremely weak, easily drowned out by contact vibration noise and electromagnetic interference caused by motor operation, resulting in a low signal-to-noise ratio of the leakage signal, making it difficult to reliably extract leakage characteristics from background noise; Second, the change in internal humidity of the motor is affected not only by the evaporation rate of the leaked water, but also by the internal temperature fluctuations caused by motor start-up and shutdown, and the condensation and re-evaporation process of water vapor in the air. The overlapping of these factors results in the humidity change curve being highly similar to the condensation-drying cycle curve under normal operating conditions in the initial stage of leakage. A single humidity value at a given moment cannot reliably distinguish between leakage and condensation. Thirdly, the aging degree of the seal and its operating time are not simply linearly related; they also depend on the difference in accelerated aging rates of the seal material under different temperatures and media immersion conditions. If only the operating time is used as a uniform time variable input, without considering the nonlinear weighting effect of historical temperature history and media contact history on the aging rate, the assessment of the remaining life of the seal and the leakage risk level will be overly optimistic under long-term low-load operating conditions and overly lagging under frequent high-temperature start-stop conditions, leading to inaccurate triggering of graded warnings. To solve the above problems, this embodiment provides a specific implementation method that inputs the leakage signal from the sealing joint surface, the internal humidity of the motor, and the operating time into a pre-constructed seal failure prediction model to output the aging degree of the seal and the leakage risk level, and triggers graded warnings. Specifically:

[0126] E101. Acquire the leakage signal of the sealing joint surface, preprocess the leakage signal of the sealing joint surface, and extract the effective leakage feature value. In this embodiment, the leakage signal of the sealing joint surface is collected by a conductivity-type leakage sensor installed below the sealing joint surface. The conductivity-type leakage sensor outputs a reference voltage of 0 volts in a dry state. When leakage occurs at the sealing joint surface and the leaking liquid contacts the sensor electrode, the output voltage increases with the increase of liquid conductivity and liquid coverage area. The preprocessing includes: collecting the raw voltage signal output by the leakage sensor at a preset first sampling frequency, forming a sliding window with the most recently collected Q raw voltage sample values, where Q is the integer value of the product of the first sampling frequency and 1 second; calculating when The arithmetic mean of the Q samples within the previous sliding window is used as the average voltage value Vavg of the current window. The standard deviation of the Q samples within the current sliding window is calculated as the noise intensity σn of the current window. The average voltage value Vavg of the current window is compared with the leakage signal noise floor threshold Vth_noise preset in the adjustment module. If Vavg is less than Vth_noise, the current effective leakage feature value Vleak is set to 0. If Vavg is greater than or equal to Vth_noise, the signal-to-noise ratio SNR of the current window is calculated as Vavg divided by σn. When SNR is less than the preset signal-to-noise ratio threshold SNR... When SNR is greater than or equal to SNR_th, it is determined that the voltage rise in the current window is caused by contact vibration noise or electromagnetic interference caused by motor vibration, and the current effective leakage characteristic value Vleak is still set to 0; when SNR is greater than or equal to SNR_th, it is determined that there is a real leakage signal in the current window, and Vavg minus Vth_noise is used as the current effective leakage characteristic value Vleak; wherein, the leakage signal noise floor threshold Vth_noise is determined during the motor factory test stage, after confirming that the sealing joint surface is completely dry and the motor is running at rated speed under various load conditions, by continuously recording the maximum fluctuation value of the leakage sensor output voltage. The average of the maximum fluctuation values ​​under each operating condition, multiplied by a safety factor of 1.2, is preset in the adjustment module. The signal-to-noise ratio (SNR) threshold SNR_th is set to 3, meaning that the effective signal amplitude must reach at least three times the noise intensity to be accepted. This setting is based on the following: During the motor's factory testing phase, with the sealing joint surface completely dry and the motor running at rated speed and rated load, the raw voltage signal output by the leakage sensor is continuously collected. The probability distribution characteristics of its noise intensity are statistically analyzed, and the noise floor mean and standard deviation are calculated. Simultaneously, a small amount of leakage is artificially introduced into the sealing joint surface, and the minimum resolvable signal amplitude of the leakage sensor's output voltage is recorded. According to the commonly used engineering discrimination criteria in the field of weak signal detection, when the effective signal value reaches three times the effective noise value (i.e., a signal-to-noise ratio of 3, equivalent to approximately 9.5 dB), the signal can be considered statistically significantly different from the background noise, achieving an acceptable balance between false alarm rate and false negative rate.The signal-to-noise ratio threshold is preset in the adjustment module as a constant value. During real-time operation, the presence of a real leakage signal in the window is determined only when the ratio of the average voltage value to the noise intensity within the current sliding window is greater than or equal to the threshold. This effectively suppresses the influence of contact vibration noise and electromagnetic interference caused by motor vibration on leakage determination, ensures the reliability of effective leakage feature value extraction, and provides accurate input parameters for subsequent sealing failure prediction models. In this embodiment, the preset first sampling frequency is determined as follows: During the system factory debugging phase, the time required for the output voltage of the leak sensor at the sealing joint surface to rise from the reference value to 63.2% of the steady-state value when the measured liquid contacts the electrode surface is obtained through a step test. This time is taken as the response time constant τ_leak of the leak sensor. According to the signal sampling theorem, in order to ensure complete capture of the signal change characteristics during the response process, the sampling frequency should not be less than 5 to 10 times the reciprocal of the response time constant. At the same time, combined with the cutoff frequency fc of the low-pass filter of the leak sensor signal output circuit hardware, in order to avoid aliasing, the sampling frequency should not be less than 2fc. Considering the above constraints, the larger value between the lowest sampling frequency that meets the response time constant requirement and 2fc is taken, and further rounded to a standard frequency that is convenient for digital system frequency division. This value is taken as the preset first sampling frequency and is preset in the storage unit of the adjustment module in the form of configuration parameters for real-time acquisition of the leak signal. In this embodiment, the safety factor of 1.2 is set as follows: During the motor factory test, under the condition that the sealing joint surface is completely dry and the motor is running at rated speed in sequence under various typical load conditions, the maximum fluctuation value of the output voltage of the leakage sensor is continuously recorded, and the arithmetic mean of the maximum fluctuation value under each condition is taken as the noise floor mean. Considering that in the actual operating environment, factors such as motor vibration intensity, electromagnetic interference level, and ambient temperature and humidity may cause a certain degree of random drift in the noise floor of the leakage sensor output signal, in order to achieve a balance between suppressing false alarms and ensuring detection sensitivity, in engineering practice, the noise floor mean is usually increased by 20% as a safety boundary, that is, multiplied by a safety factor of 1.2. This factor value refers to the general engineering practice of weak signal extraction in the field of industrial leakage detection. It can both encompass the random fluctuation range of noise under normal operating conditions and avoid false judgments triggered by instantaneous noise spikes, and prevent the threshold from being too high to mask the early characteristics of minor leakage, thereby ensuring the robustness and reliability of the leakage signal noise floor threshold Vth_noise in long-term operation.

[0127] E102. Obtain the internal humidity of the motor, preprocess the internal humidity of the motor, and extract the effective humidity feature value. In this embodiment, the internal humidity of the motor is collected by a humidity sensor installed on the top of the motor cavity away from the cooling circuit pipe inside the shaft. The humidity sensor outputs a relative humidity percentage signal. The preprocessing includes: collecting the relative humidity value output by the humidity sensor at a preset second sampling frequency, which is set to 1 Hz; obtaining the relative humidity time series within the previous 60 sampling periods (60 seconds) before the current time, performing linear regression on the time series to obtain the humidity change slope Slp_H, in units of relative humidity percentage per hour; and simultaneously, acquiring the current humidity feature value collected by the sensing module. The internal temperature of the front motor is used to calculate the difference between the current internal temperature of the motor and the current internal dew point temperature of the motor as the dew point margin. The current internal dew point temperature of the motor is obtained by referring to a humid air dew point temperature lookup table preset in the adjustment module based on the current internal humidity and the current internal temperature of the motor. The humidity change slope Slp_H is compared with a preset first slope threshold Slp_th1, and the dew point margin is compared with a preset dew point margin threshold Dew_th. When Slp_H is less than Slp_th1 and the dew point margin is greater than Dew_th, the current humidity change is determined to be in normal condensation. - In the drying cycle state, the effective humidity feature value Hfeature is set to 0; when Slp_H is greater than or equal to Slp_th1 and the dew point margin is less than or equal to Dew_th, it is determined that the current humidity change is caused by the continuous evaporation of seepage water, and Slp_H multiplied by the preset integration time window length Δt is used as the effective humidity feature value Hfeature; wherein, the first slope threshold Slp_th1 is set to 5% per hour, indicating that when the relative humidity rises by no more than 5% per hour, it is considered a normal condensation cycle; in this embodiment, the dew point margin threshold Dew_th is set to 5℃ is used to characterize the proximity of the motor's internal temperature to the current dew point temperature. When this difference is less than or equal to 5℃, it indicates that the air inside the motor is nearly saturated. At this time, if there is a trace of water leakage, its evaporation will significantly increase the relative humidity and form a continuous upward trend in humidity. At the same time, small temperature fluctuations may also trigger condensation. Therefore, this threshold is used as a criterion to distinguish between normal condensation-drying cycles and abnormal humidity increases caused by leakage. It is determined comprehensively based on the thermodynamic characteristics of humid air and the temperature fluctuation range under typical operating conditions inside the motor, and is preset in the adjustment module. The integral time window length Δt is set to 1 hour. In this embodiment, the preset basis for setting the second sampling frequency to 1 Hz is as follows: During the system factory debugging stage, the response time constant τ_humi of the humidity sensor is first obtained through a step response test, that is, when the humidity inside the motor undergoes a step change, the time required for the humidity sensor output to rise from the initial value to the steady-state value of 63.2%. According to the sampling theorem, in order to ensure complete capture of the dynamic change process of the humidity signal, the sampling frequency should not be less than 5 to 10 times the reciprocal of the response time constant.Meanwhile, considering that the humidity changes inside the motor are affected by leakage evaporation or condensation-drying cycles, and its typical timescale is on the order of minutes, much slower than rapidly changing parameters such as temperature or pressure, an excessively high sampling frequency will introduce a large amount of redundant data and increase the computational load on the adjustment module, while an excessively low sampling frequency will fail to capture the starting inflection point of abnormal humidity changes in a timely manner. Taking into account the actual response time of the humidity sensor, the physical timescale of humidity changes inside the motor, and the requirement for data continuity in the calculation of the humidity change slope in subsequent preprocessing, a sampling frequency of 1 time per second (i.e., 1 Hz) is selected as the second sampling frequency. This frequency is pre-set as a configuration parameter in the storage unit of the adjustment module for real-time acquisition of the relative humidity value inside the motor.

[0128] E103. Convert the actual running time into a weighted equivalent aging time, and use the weighted equivalent aging time as a characterization of the aging degree of the seal. In this embodiment, the weighted equivalent aging time is calculated as follows: The internal temperature of the motor and the inlet temperature of the shaft cooling circuit obtained by the sensing module are collected in a preset third sampling period, which is set to 1 minute. The higher of the internal temperature of the motor and the inlet temperature of the shaft cooling circuit collected in each sampling period is taken as the working temperature T_seal of the seal for that sampling period. Based on the sealing material temperature accelerated aging factor mapping table preset in the adjustment module, the temperature accelerated aging factor AF_T corresponding to that sampling period is obtained by looking up the table or using linear interpolation, with T_seal as the index. The temperature accelerated aging factor mapping table is established by fitting the Arrhenius equation to the accelerated aging test data of the sealing material at multiple constant temperatures, with temperature as the independent variable and the ratio of the aging rate at that temperature to the aging rate at the reference temperature as the dependent variable. Simultaneously, based on the sealing material medium contact accelerated aging factor mapping table preset in the adjustment module, the medium contact accelerated aging factor AF_M corresponding to that sampling period is obtained by looking up the table or using linear interpolation, with the inlet temperature of the shaft cooling circuit as the medium temperature index. The sub-mapping table is established based on immersion aging test data of the sealing material under different temperature combinations of conveying medium. The temperature-accelerated aging factor AF_T is multiplied by the medium-contact accelerated aging factor AF_M to obtain the comprehensive accelerated aging factor AF for that sampling period. The actual duration of that sampling period is multiplied by the comprehensive accelerated aging factor AF to obtain the equivalent aging time increment for that sampling period. The equivalent aging time increments of all sampling periods from the time of sealing replacement or equipment commissioning to the current time are accumulated to obtain the weighted equivalent aging time T_eq. T_eq is compared with the pre-set sealing design reference life T_design in the adjustment module. When T_eq is less than 50% of T_design, the sealing aging degree is determined to be mild aging; when T_eq is between 50% and 85% of T_design, the sealing aging degree is determined to be moderate aging; when T_eq is greater than 85% of T_design, the sealing aging degree is determined to be severe aging. The sealing design reference life T_design is the rated service life provided by the sealing manufacturer under reference temperature and clean medium conditions.

[0129] E104. Input the effective leakage feature value Vleak obtained from E101, the effective humidity feature value Hfeature obtained from E102, and the weighted equivalent aging time T_eq and the aging degree of the seal obtained from E103 into the pre-constructed seal failure prediction model, and output the leakage risk level. In this embodiment, the seal failure prediction model is a rule-based decision tree model. The specific decision rules are as follows: Rule 1: If Vleak equals 0 and Hfeature equals 0, then regardless of the aging degree of the seal, the leakage risk level is determined to be level 0, i.e., no leakage risk; Rule 2: If Vleak equals 0 and Hfeature is greater than 0, then the leakage risk level is determined according to the aging degree of the seal: when the aging degree of the seal is light, the leakage risk level is determined to be level 1, i.e., low risk; when the aging degree of the seal is moderate, the leakage risk level is determined to be level 2, i.e., medium risk; when the aging degree of the seal is heavy, the leakage risk level is determined to be level 3, i.e., high risk; Rule 3: If Vleak is greater than 0 and Hfeature equals 0, then the leakage risk level is determined to be level 1, i.e., high risk; ...2, i.e., medium risk; when the aging degree of the seal is heavy If Vleak is 0, then Vleak is compared with the preset first leakage voltage threshold Vth1: when Vleak is less than Vth1, the leakage risk level is determined to be level 1; when Vleak is greater than or equal to Vth1 and less than the second leakage voltage threshold Vth2, the leakage risk level is determined to be level 2; when Vleak is greater than or equal to Vth2, the leakage risk level is determined to be level 3; rule 4, if Vleak is greater than 0 and Hfeature is greater than 0, then the leakage risk level is directly determined to be level 3, i.e., high risk; where the first leakage voltage threshold Vth1 is set to 0.5 volts and the second leakage voltage threshold Vth2 is set to 2.0 volts, both are determined based on the calibration test results of the leakage sensor under simulated leakage rates and preset in the adjustment module.

[0130] E105. The output leakage risk level triggers a corresponding graded warning signal when the leakage risk level reaches the specified level. In this embodiment, when the leakage risk level is level 1, a blue warning signal is triggered, pushing an information package containing the current timestamp, leakage risk level, effective leakage characteristic value, effective humidity characteristic value, and seal aging degree to the cloud platform, prompting operators to include the water pump in the monthly inspection plan for key monitoring; when the leakage risk level is level 2, a yellow warning signal is triggered, pushing the above information package to the cloud platform with the additional recommendation to arrange for a one-week shutdown to check the seal status; when the leakage risk level is level 3, a red warning signal is triggered, pushing the above information package to the cloud platform with the additional recommendation to immediately shut down the pump and replace the seal, while simultaneously issuing an alarm indication through the local audible and visual alarm. If the leakage risk level is level 0, no warning signal is triggered, and only the current assessment result is recorded in the local log.

[0131] In terms of leakage signal processing, this embodiment uses a sliding window sampling method to sample the original voltage signal of the leakage sensor and calculates the signal-to-noise ratio (SNR) within the window. A SNR threshold is used to distinguish between the actual leakage signal and contact vibration noise and electromagnetic interference caused by motor vibration. Only when the SNR meets the requirements is a valid leakage feature value extracted. This solves the problem that the signal amplitude is weak and easily submerged by background noise during the micro-leakage stage, making reliable extraction of leakage features impossible. In terms of humidity signal processing, a linear regression is performed on the relative humidity time series inside the motor to obtain the humidity change slope. Simultaneously, the difference between the current internal temperature of the motor and the dew point temperature is used as the dew point margin. The humidity change slope and dew point margin are jointly used to distinguish between normal condensation-drying cycles and humidity changes caused by continuous evaporation of leaked water. This solves the problem that the humidity change curve in the early stage of leakage is highly similar to that of a normal condensation cycle, and a single humidity value cannot reliably distinguish between leakage and condensation. In terms of seal aging assessment, seal aging data is collected at minute intervals. Based on the operating temperature of the component, a temperature-accelerated aging factor mapping table and a media contact-accelerated aging factor mapping table were established using the Arrhenius equation to obtain the comprehensive accelerated aging factor for each cycle. The actual operating time was weighted and accumulated cycle by cycle to obtain the weighted equivalent aging time. The aging degree of the seal was determined by the ratio of this weighted equivalent aging time to the design baseline life. This solves the problem of evaluation bias caused by the fact that the operating time is only used as a uniform time variable input in the existing technology, which cannot reflect the difference in accelerated aging rate under different temperature and media conditions. Finally, by inputting the effective leakage characteristic value, the effective humidity characteristic value, and the aging degree of the seal into a rule-based decision tree model for multi-factor fusion judgment, the leakage risk level is output and graded early warning is triggered according to the level. This realizes the leap from passive fault alarm to active degradation trend prediction, enabling operators to obtain graded early warning information in the early stage of seal failure and take corresponding maintenance measures to avoid motor water ingress damage caused by sudden seal failure.

[0132] Example 5

[0133] It should be noted that if the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet is directly used as the first temperature difference, and the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet is used as the second temperature difference, and the current actual heat dissipation efficiency is directly calculated from the second temperature difference and the current cooling contribution of the external forced air cooling circuit, and then the current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency to obtain the heat dissipation efficiency attenuation rate, and then the degree of external contamination of the water pump is determined by combining the first temperature difference, the heat dissipation efficiency attenuation rate, and the cumulative running time, three lower-level technical problems will be encountered: First, the surface temperature of the casing is directly affected by the balance between the heat generated inside the motor and the heat removed by the internal cooling circuit. When the motor load fluctuates or the cooling capacity of the internal cooling circuit changes, even if the degree of external heat dissipation fin contamination does not change, the surface temperature of the casing will still deviate significantly, causing the change in the first temperature difference to not be uniquely mapped to the change in the degree of external contamination; Second, the second temperature difference calculated from the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet, when the current cooling contribution of the external forced air cooling circuit is... When the contribution of cooling to heat dissipation decreases due to increased airflow resistance or filter blockage, the reduced cooling airflow leads to an increase in the temperature of the heat dissipation fins and a larger second temperature difference. However, this increase in the second temperature difference is exactly opposite to the decrease in heat dissipation efficiency; that is, the more severe the pollution, the larger the second temperature difference. If the second temperature difference is directly combined with the current cooling contribution to calculate the actual heat dissipation efficiency, the increase in the second temperature difference caused by increased pollution will partially offset the impact of the decrease in cooling contribution on heat dissipation efficiency, thus significantly reducing the sensitivity of the calculated heat dissipation efficiency decrease in terms of the degree of pollution. Thirdly, the cumulative running time only reflects the total running time of the water pump since its commissioning and cannot distinguish the differences in external pollution conditions such as dust concentration and humidity in different running periods. Directly using the cumulative running time as an auxiliary criterion for the degree of pollution cannot provide a differentiated assessment of the degree of pollution in two scenarios: long-term operation in a clean environment and short-term operation in a dusty environment. This results in the online cleaning triggering timing being seriously delayed compared to the actual degree of pollution accumulation or being triggered too early, leading to a waste of cleaning resources. To solve the above problems, this embodiment provides a specific implementation method for determining the degree of external pollution of the water pump and triggering online cleaning, specifically:

[0134] F101. Obtain the current casing surface temperature T_shell, current heat sink fin temperature T_fin, ambient temperature at the far end of the air inlet T_amb_remote, current cooling contribution of the external forced air cooling circuit P_fan_current, current total heat load P_heat_total determined by A106, and current cooling contribution of the in-shaft cooling circuit P_pump_current determined by B107 from the sensing module.

[0135] F102. Equivalent thermal resistance of the computer casing, using the equivalent thermal resistance of the casing as the first pollution characterization quantity. In this embodiment, the equivalent thermal resistance of the casing R_shell is calculated as follows: subtract the current cooling contribution of the internal cooling circuit P_pump_current from the current total heat load P_heat_total to obtain the remaining heat to be dissipated outward through the casing P_shell = P_heat_total - P_pump_current; use the difference between the current casing surface temperature T_shell and the ambient temperature T_amb_remote at the far end of the air inlet as the effective heat transfer temperature difference of the casing ΔT_shell = T_shell - T_amb_remote; if ΔT_shell ≤ 0, then the equivalent thermal resistance of the casing R_shell is set to the preset maximum value R_max, and a thermal balance abnormality warning signal is triggered; if ΔT_shell > 0, then the effective heat transfer temperature difference of the casing ΔT_shell is divided by the remaining heat P_shell to obtain the equivalent thermal resistance of the casing R_shell = ΔT_shell / P_shell, in degrees Celsius per kilowatt; the calculated equivalent thermal resistance R_shell of the casing is compared with the pre-set clean-state reference thermal resistance R_shell_0 in the adjustment module to obtain the casing thermal resistance increase factor K_shell = R_shell / R_shell_0; the clean-state reference thermal resistance R_shell_0 is the equivalent thermal resistance value of the casing calculated in the same way as above during the factory test stage, under the conditions that the outer surface of the casing and the heat dissipation fins are in a factory clean state and the external forced air cooling circuit is running at the rated speed, and it is pre-set in the adjustment module. In this embodiment, the preset maximum value R_max is set as follows: during the system factory debugging stage, the maximum legal value that the floating-point variable used to store the equivalent thermal resistance of the casing in the adjustment module can represent under its data type specification is obtained, the maximum legal value is multiplied by a safety derating factor of 0.8 and rounded down to the nearest integer as the preset maximum value R_max, and it is pre-set in the storage unit of the adjustment module as a configuration parameter. This value ensures that it significantly exceeds the actual fluctuation range of the equivalent thermal resistance of the casing under normal operating conditions (typically 0.1 to 10 degrees Celsius per kilowatt), clearly indicating an abnormal thermal balance state when it appears in the calculation results. It also retains sufficient numerical margin to avoid data overflow in subsequent multiplication or accumulation operations. When ΔT_shell is determined to be less than or equal to zero during real-time calculation, R_shell is directly assigned the value R_max, simultaneously triggering a thermal balance anomaly warning signal. This makes the maximum value a traceable anomaly marker rather than valid data for subsequent pollution level scoring.

[0136] F103. Calculate the normalized heat dissipation efficiency attenuation rate, and use the normalized heat dissipation efficiency attenuation rate as the second pollution characterization quantity. In this embodiment, the normalized heat dissipation efficiency attenuation rate is calculated as follows: obtain the current cooling contribution P_fan_current of the external forced air cooling circuit determined by B207; obtain the effective heat transfer temperature difference ΔT_fan_eff determined by B206, that is, the difference between the average temperature of the heat dissipation fins T_fin_avg and the ambient temperature T_amb_remote at the far end of the air inlet, ΔT_fan_eff = T_fin_avg - T_amb_remote; the average temperature of the heat dissipation fins T_fin_avg is obtained in the same way as in B203, by calculating the arithmetic mean of the temperature values ​​collected from 9 temperature measuring points on the surface of the heat dissipation fins; divide the current cooling contribution P_fan_current of the external forced air cooling circuit by the effective heat transfer temperature difference ΔT_fan_eff to obtain the current actual heat transfer coefficient UA_current = P_fan_current / ΔT_fan_eff, in kilowatts per degree Celsius; obtain the pre-set clean-state reference heat transfer coefficient UA_0 in the adjustment module. The clean-state reference heat transfer coefficient UA_0 is calculated in the same manner as above during the factory test phase, under the reference operating conditions where the inlet filter and heat dissipation fins are in a factory clean state, the cooling fan blades operate at rated speed, and the total heat load is stable at 50% of the rated load; divide the current actual heat transfer coefficient UA_current by the clean-state reference heat transfer coefficient UA_0 to obtain the current heat transfer coefficient retention rate η_UA = UA_current / UA_0; subtract the current heat transfer coefficient retention rate η_UA from 1 to obtain the normalized heat dissipation efficiency attenuation amplitude Δη = 1 - η_UA, the normalized heat dissipation efficiency attenuation range Δη is 0 to 1, where 0 indicates that the heat dissipation efficiency has not decreased and 1 indicates that the heat dissipation efficiency has been completely lost. By introducing the heat transfer coefficient as a normalized index, the coupling relationship between the second temperature difference and the cooling contribution is transformed into a scalar that is directly comparable to the clean state benchmark, thus eliminating the offsetting effect of the increase in the second temperature difference on the attenuation of heat dissipation efficiency when pollution intensifies.

[0137] F104. Calculate the environmental weighted cumulative pollution duration, and use the environmental weighted cumulative pollution duration as the third pollution characterization quantity. In this embodiment, the environmental weighted cumulative pollution duration is calculated as follows: the ambient temperature T_amb_remote at the far end of the air inlet obtained by the sensing module and the pressure difference ΔP_filter across the air inlet filter obtained by B202 are collected at a preset pollution assessment cycle. The pollution assessment cycle is set to 10 minutes, and the setting is based on the following: during the system factory commissioning stage, the balance between the physical time scale of the filter clogging process, the typical cycle of environmental dust concentration fluctuation, and the calculation load of the adjustment module is comprehensively considered. Specifically, the accumulation of dust on the filter surface is a slow, gradual process. The increase in clogging pressure differential does not change abruptly within minutes. An excessively short assessment cycle would lead to frequent readings of pressure differential data and the execution of weighted summation calculations, increasing the computational overhead of the adjustment module without providing more valuable pollution trend information. Meanwhile, although the concentration of dust in the external environment fluctuates instantaneously with changes in operating conditions or wind direction, the statistically significant effective period for its actual impact on dust accumulation on heat sink fins is typically only a few minutes to tens of minutes. Therefore, using 10 minutes as the pollution assessment cycle allows for sufficient time granularity to capture the transition from clean to dusty environmental conditions, ensuring that the switching of environmental weighting coefficients reflects changes in actual pollution conditions in a timely manner, while avoiding data redundancy and wasted computational resources due to overly frequent sampling.This cycle value is pre-set as a configuration parameter in the storage unit of the adjustment module at the factory. During real-time operation, it uses this fixed cycle as the beat to synchronously collect the ambient temperature at the far end of the air inlet and the pressure difference value across the filter, which is used for updating the environmental weighted cumulative pollution duration calculation. The pressure difference value ΔP_filter across the filter at the air inlet collected in each pollution assessment cycle is compared with the filter cleanliness benchmark pressure difference value ΔP_filter_0 pre-set in the adjustment module. When ΔP_filter - ΔP_filter_0 < the preset pollution pressure difference threshold ΔP_th, it is determined that the environment is in a clean state during the pollution assessment cycle, and the environmental weight coefficient W_env for the pollution assessment cycle is set to the benchmark value of 1.0. When ΔP_filter - ΔP_filter_0 ≥ ΔP_th, it is determined that the environment is in a dusty state during the pollution assessment cycle. Based on the ambient temperature at the far end of the air inlet T_amb_remote of the cycle, the dust adhesion enhancement factor K_dust is obtained from the temperature-dust adhesion enhancement factor mapping table pre-set in the adjustment module by lookup or interpolation. The environmental weight coefficient W_env is then set. = K_dust, where K_dust > 1.0, indicating that dust is more likely to adhere to the fin surface under high temperature conditions; multiply the actual duration of each pollution assessment cycle by the environmental weight coefficient W_env for that cycle to obtain the weighted pollution duration for that cycle; sum the weighted pollution durations of all pollution assessment cycles from the time the equipment was put into operation or the time of the most recent online cleaning to the current time to obtain the environmental weighted cumulative pollution duration T_dirt_eq; the pollution pressure difference threshold ΔP_th = 0.5 × In this embodiment, the pollution differential pressure threshold ΔP_th is set to 0.5 times the baseline differential pressure value ΔP_filter_0 for the clean state of the filter. The basis for this setting is as follows: During the system's factory commissioning phase, a bench test was conducted on the inlet filter to gradually clog it, recording the entire process curve of the filter clogging differential pressure increment gradually increasing from zero. The test results show that when the filter clogging differential pressure increment reaches 0.5 times the baseline differential pressure value for the clean state, an identifiable initial dust layer has formed on the filter surface, and the inlet begins to transition from a clean environment to a dusty environment. At this time, the switching of the environmental weight coefficient can promptly reflect changes in external pollution conditions. This value avoids both excessively low thresholds that lead to frequent erroneous switching of environmental states caused by sensor noise or instantaneous wind pressure fluctuations, and excessively high thresholds that lead to lag in the identification of slightly dusty environments, thus achieving a balance between the sensitivity and stability of distinguishing between clean and dusty environments. The threshold is preset as a constant value in the storage unit of the adjustment module at the factory; in this embodiment, the temperature-dust adhesion enhancement factor mapping table is established under laboratory conditions by simulating a dusty environment and controlling different temperature conditions.The specific method is as follows: Water pump heat sink fin material samples are placed in a reference ambient temperature of 20℃ and several constant-temperature environments above 20℃. Standard test dust with the same concentration and particle size distribution is introduced into each constant-temperature environment. Within the same exposure time, the average dust accumulation per unit area of ​​the fin samples is measured under each temperature condition. Using the average dust accumulation per unit area under the 20℃ reference temperature as a reference value, the average dust accumulation per unit area measured under each temperature condition is divided by this reference value. The resulting ratio is the dust adhesion enhancement factor K_dust corresponding to that temperature. This factor quantitatively characterizes the relative degree to which increased ambient temperature leads to enhanced dust adhesion and accelerated dust accumulation rate on the fin surface. Each test temperature point and its corresponding dust adhesion enhancement factor are stored in a mapping table, pre-installed in the adjustment module. When the environment is determined to be dusty, the corresponding K_dust value is obtained by looking up the table based on the real-time collected ambient temperature at the far end of the air inlet, and used as an environmental weighting coefficient in the calculation of the weighted cumulative pollution duration.

[0138] Input F105, the casing thermal resistance increase factor K_shell obtained from F102, the normalized heat dissipation efficiency attenuation magnitude Δη obtained from F103, and the environmental weighted cumulative pollution duration T_dirt_eq obtained from F104 into the pre-built comprehensive pollution degree judgment model, and output the external pollution degree level of the water pump. In this embodiment, the comprehensive pollution degree judgment model is a multi-factor fusion model based on weighted scoring. Specifically, the judgment method is as follows: the increase factor of the casing thermal resistance K_shell is mapped to the casing pollution score Score_shell, with the following mapping rules: when K_shell < 1.2, Score_shell = 0; when 1.2 ≤ K_shell ≤ 1.5, Score_shell = 30 × (K_shell - 1.2) / 0.3; when 1.5 < K_shell ≤ 2.0, Score_shell = 30 + 20 × (K_shell - 1.5) / 0.5; when K_shell > 2.0, Score_shell = 50. The normalized heat dissipation efficiency attenuation magnitude Δη is mapped to the fin pollution score Score_fin, with the following mapping rules: when Δη < 0.1, Score_fin = 0; when 0.1 ≤ Δη ≤ 0.3, Score_fin = 40 × (Δη - 0.1) / 0.2; When 0.3 < Δη ≤ 0.5, Score_fin = 40 + 30 × (Δη - 0.3) / 0.2; When Δη > 0.5, Score_fin = 70; Map the environmental weighted cumulative pollution duration T_dirt_eq to the duration pollution score Score_time. The mapping rule is as follows: Compare T_dirt_eq with the pre-set design cleaning interval T_clean_design in the adjustment module. When T_dirt_eq < 0.5 × T_clean_design, Score_time = 0; When 0.5 × T_clean_design ≤ T_dirt_eq ≤ T_clean_design, Score_time = 20 × (T_dirt_eq - 0.5 × T_clean_design) / (0.2);5 × T_clean_design); When T_dirt_eq > T_clean_design, Score_time = 20; The casing contamination score Score_shell, the fin contamination score Score_fin, and the duration contamination score Score_time are directly added together to obtain the comprehensive contamination score Score_total = Score_shell + Score_fin + Score_time; When Score_total < 30, the contamination level is determined to be level 0, i.e., clean; when 30 ≤ Score_total < 60, the contamination level is determined to be level 1, i.e., lightly contaminated; when 60 ≤ Score_total < 85, the contamination level is determined to be level 2, i.e., moderately contaminated; when Score_total ≥ 85, the contamination level is determined to be level 3, i.e., heavily contaminated; The designed cleaning interval T_clean_design is set by the operator and preset in the adjustment module based on the annual average dust concentration of the expected operating environment of the water pump.

[0139] F106. Compare the pollution level output by F105 with the preset cleaning trigger threshold. When the pollution level reaches the preset cleaning trigger threshold, start the online cleaning program and control the start / stop of the cleaning device, the cleaning water pressure, and the cleaning time. In this embodiment, the preset cleaning trigger threshold is set to level 2, meaning that cleaning is triggered when the pollution level is determined to be moderate or heavy. The cleaning water pressure is set as follows: when the pollution level is level 2, the cleaning water pressure is set to 70% of the rated cleaning water pressure; when the pollution level is level 3, the cleaning water pressure is set to 100% of the rated cleaning water pressure. The cleaning duration is set as follows: when the pollution level is level 2, the cleaning duration is set to 1 times the preset standard cleaning duration; when the pollution level is level 3, the cleaning duration is set to 1.5 times the preset standard cleaning duration. The rated cleaning water pressure and the preset standard cleaning duration are provided by the cleaning device manufacturer and preset in the adjustment module. After the online cleaning program is started, the adjustment module first outputs a cleaning device start command and controls the cleaning water pressure to reach the set value, continues to run until the set cleaning duration, and then outputs a cleaning device stop command. After cleaning is completed, the environmental weighted cumulative pollution duration T_dirt_eq in F104 is reset to 0, and the accumulation starts again. If the pollution level is 0 or 1, cleaning will not be triggered; only the current comprehensive pollution score and various pollution characterization quantities will be recorded in the local log.

[0140] In terms of casing contamination characterization, the remaining heat to be dissipated through the casing is obtained by subtracting the current cooling contribution of the internal cooling circuit from the current total heat load. The equivalent thermal resistance of the casing is then calculated by dividing the effective heat transfer temperature difference of the casing by this remaining heat. The increase in casing thermal resistance is then compared with the baseline thermal resistance in a clean state, and this increase is used as the first contamination characterization factor. This eliminates the interference of casing surface temperature deviation caused by motor load fluctuations or changes in the cooling capacity of the internal cooling circuit on the contamination determination, making the characterization of the contamination degree on the casing side independent of the internal thermal balance of the motor. Regarding the characterization of fin contamination, the heat transfer coefficient is introduced as a normalized index. The current actual heat transfer coefficient is obtained by dividing the current cooling contribution of the external forced air cooling loop by the effective heat transfer temperature difference. The current actual heat transfer coefficient is then compared with the clean-state baseline heat transfer coefficient to obtain the heat transfer coefficient retention rate. The complement of the heat transfer coefficient retention rate is then used as the normalized heat dissipation efficiency attenuation rate. This transforms the mutually coupled and oppositely changing relationship between the second temperature difference and the cooling contribution into a normalized scalar directly comparable to the clean-state baseline, eliminating the influence of contamination additives. The increased temperature difference during peak hours offsets the decrease in heat dissipation efficiency, significantly improving the sensitivity of the decrease in heat dissipation efficiency to the degree of contamination. Regarding cumulative contamination duration, the system differentiates between clean and dusty environments by using the increase in filter clogging pressure difference as a criterion. In dusty environments, the system assigns differentiated environmental weight coefficients to each contamination assessment cycle based on the dust adhesion enhancement factor obtained from the remote ambient temperature, and then weights and accumulates the cumulative contamination duration. This allows the cumulative contamination duration to distinguish the differences in contamination rates under different environmental conditions, avoiding the problem that ordinary cumulative runtime cannot provide discriminative assessment results in clean and dusty environments. Finally, by inputting the three independent indicators representing the degree of contamination into a multi-factor fusion model based on weighted scoring for comprehensive judgment, and matching differentiated cleaning water pressure and cleaning duration according to the contamination level, the system achieves accuracy in contamination degree judgment, precision in cleaning triggering timing, and matching of cleaning parameters with the degree of contamination. This avoids problems of untimely or excessive cleaning due to misjudgment, improving the intelligence level of online cleaning and the operational reliability of the water pump throughout its entire lifecycle.

[0141] Example 6

[0142] Another embodiment of the present invention provides: an adaptive control method for a fully sealed dual-cooled waterproof pump, comprising:

[0143] The sensor signals deployed at key locations of the water pump are collected in real time. The sensor signals include at least: inlet temperature of the internal cooling circuit, outlet temperature of the internal cooling circuit, flow rate of the internal cooling circuit, temperature of the motor stator winding, temperature of the front bearing, temperature of the rear bearing, surface temperature of the housing, temperature of the heat sink fins, leakage signal of the sealing joint surface, internal humidity of the motor, ambient temperature at the far end of the air inlet, motor current, motor voltage, water pump speed, water pump inlet pressure and water pump outlet pressure.

[0144] Based on the sensor signals, the motor input power is calculated using the motor current and motor voltage, and the total heat load is determined using the motor stator winding temperature, front bearing temperature, rear bearing temperature, and motor input power.

[0145] Meanwhile, the current cooling contribution of the internal cooling circuit is calculated from the difference between the inlet and outlet temperatures of the internal cooling circuit and the flow rate of the internal cooling circuit. The current cooling contribution of the external forced air cooling circuit is calculated from the difference between the ambient temperature at the far end of the air inlet and the temperature of the heat dissipation fins and the speed of the cooling fan blades.

[0146] With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated. At the same time, they are converted into corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands, and the corresponding in-shaft cooling flow rate adjustment commands and / or cooling fan speed adjustment commands are executed.

[0147] Based on the sensor signals, the leakage signal of the sealing joint surface, the internal humidity of the motor and the running time are input into the pre-built sealing failure prediction model, and the assessment results characterizing the aging degree of the seal and the leakage risk level are output. The corresponding graded early warning signal is triggered according to the leakage risk level.

[0148] Based on the sensor signal, a first temperature difference is calculated from the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet; a second temperature difference is calculated from the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet; and the current actual heat dissipation efficiency is calculated from the second temperature difference and the current cooling contribution of the external forced air cooling circuit. In an embodiment of the present invention, the current actual heat dissipation efficiency is characterized by the heat transfer coefficient (i.e., the heat dissipation power per unit temperature difference), specifically:

[0149] Current actual heat dissipation efficiency = Current cooling contribution of external forced air cooling circuit ÷ Second temperature difference;

[0150] The second temperature difference is determined by the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet (i.e., the effective heat transfer temperature difference ΔT_fan_eff). Specifically, step F103 of embodiment 5 clarifies this: the current cooling contribution P_fan_current of the external forced air cooling circuit is obtained, and the effective heat transfer temperature difference ΔT_fan_eff, calculated from the difference between the average temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet, is obtained. The quotient of these two values ​​yields the current actual heat transfer coefficient UA_current = P_fan_current / ΔT_fan_eff. This heat transfer coefficient is a direct indicator of heat dissipation efficiency. The larger this value, the more heat is dissipated by the fins under the same temperature difference, indicating higher heat dissipation efficiency, slower equipment temperature rise, and more stable operation. By quantifying heat dissipation efficiency as a heat transfer coefficient, the judgment interference caused by the increase in temperature difference when pollution intensifies can be eliminated, more accurately reflecting the true attenuation of the fins' heat dissipation capacity.

[0151] The current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency corresponding to the clean state under the same working conditions to obtain the heat dissipation efficiency attenuation range. Combining the first temperature difference, the heat dissipation efficiency attenuation range and the cumulative running time, the degree of external pollution of the water pump is determined. When the degree of pollution reaches the preset cleaning trigger threshold, the online cleaning program is started to control the start and stop of the cleaning device, the cleaning water pressure and the cleaning time.

[0152] Example 7

[0153] An electronic device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement a fully sealed, dual-cooled, waterproof pump adaptive control system.

[0154] A computer-readable storage medium having computer instructions stored thereon, which, when executed, execute a fully sealed, dual-cooled, waterproof pump adaptive control system.

[0155] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the claims. All of these variations are within the scope of protection of the present invention.

[0156] If the technical solution disclosed herein involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

Claims

1. A fully sealed, dual-cooled, waterproof pump adaptive control system, characterized in that, The fully sealed dual-cooled waterproof pump integrates an internal shaft cooling circuit and an external forced air cooling circuit. The system includes: The sensing module is configured to: collect sensor signals deployed at key locations of the water pump in real time; the sensor signals include: inlet temperature of the internal cooling circuit, outlet temperature of the internal cooling circuit, flow rate of the internal cooling circuit, motor stator winding temperature, front bearing temperature, rear bearing temperature, housing surface temperature, heat sink fin temperature, leakage signal of the sealing joint surface, internal humidity of the motor, ambient temperature at the far end of the air inlet, motor current, motor voltage, water pump speed, water pump inlet pressure, and water pump outlet pressure; The adjustment module is configured to: calculate the motor input power based on the sensor signal using the motor current and motor voltage, and determine the total heat load using the motor stator winding temperature, front bearing temperature, rear bearing temperature, and motor input power; Determining the total heat load includes: The pump shaft power under the current operating condition is determined by the pump speed and the pre-determined pump head-speed-power mapping relationship, and the difference between the motor input power and the pump shaft power is taken as the current total power loss of the motor. The stator copper loss component in the current total power loss is corrected based on the temperature coefficient of the copper resistance of the motor stator winding, and the copper loss correction amount is obtained. The stator copper loss component is calculated from the motor current and the room temperature resistance of the stator winding. The front bearing temperature is compared with the normal temperature reference value of the front bearing under the current speed and load conditions to obtain the front bearing temperature rise deviation. The rear bearing temperature is compared with the normal temperature reference value of the rear bearing under the current speed and load conditions to obtain the rear bearing temperature rise deviation. The temperature rise deviation of the front bearing and the temperature rise deviation of the rear bearing are input into a pre-established bearing friction loss increment mapping relationship to obtain the bearing friction loss increment. The sum of the copper loss correction and the bearing friction loss increment is used as the loss correction, which is then added to the current total loss power to obtain the total heat load. Meanwhile, the current cooling contribution of the internal cooling circuit is calculated from the difference between the inlet and outlet temperatures of the internal cooling circuit and the flow rate of the internal cooling circuit. The current cooling contribution of the external forced air cooling circuit is calculated from the difference between the ambient temperature at the far end of the air inlet and the temperature of the heat dissipation fins and the speed of the cooling fan blades. The current cooling contribution of the cooling circuit within the computing axis includes: The average value of the inlet temperature and the outlet temperature is taken as the qualitative temperature of the cooling medium. Based on the qualitative temperature, the specific heat capacity and density of the cooling medium at the current temperature are obtained from the cooling medium property parameter mapping table preset in the adjustment module. At the same time, it is determined whether the flow rate of the cooling circuit inside the shaft is lower than the preset low flow rate threshold. If it is lower than the low flow rate threshold, the flow rate is corrected by a preset low flow rate correction coefficient, and the corrected flow rate is used as the flow rate value to be calculated. Otherwise, the collected flow rate value is used directly. The theoretical cooling contribution of the internal cooling circuit is calculated as the product of the flow rate, density, specific heat capacity, and the difference between the outlet temperature and the inlet temperature. The estimated heat flow rate transferred by the motor side cooling medium is obtained by calculating the difference between the motor stator winding temperature, front bearing temperature, and rear bearing temperature and the inlet temperature, multiplying each difference by its respective preset equivalent heat transfer coefficient, and then summing them. The ratio of the estimated heat flow to the theoretical cooling contribution is used as the thermal resistance correction coefficient under the current operating conditions. When the ratio is greater than 1, the theoretical cooling contribution is divided by the thermal resistance correction coefficient, and the result is the corrected current cooling contribution of the in-shaft cooling circuit. The calculation of the current cooling contribution of the external forced air cooling circuit includes: The cooling fan speed is obtained, and the current cooling fan speed is converted into the theoretical cooling air volume based on the mapping relationship between the fan speed and the theoretical air volume preset in the adjustment module. The pressure difference value on both sides of the air inlet filter is obtained, and the pressure difference value is compared with the reference pressure difference value of the filter in the clean state preset in the adjustment module to obtain the filter blockage pressure difference increment; at the same time, the fin temperature values ​​collected by multiple preset temperature measuring points on the surface of the heat dissipation fins are obtained, and the arithmetic mean of the multiple fin temperature values ​​is taken as the fin average temperature. Based on the filter clogging pressure difference increment, the airflow attenuation coefficient under the current operating condition is obtained from the mapping relationship between the filter clogging pressure difference increment and the airflow attenuation coefficient preset in the adjustment module; the theoretical cooling airflow is multiplied by the airflow attenuation coefficient to obtain the effective cooling airflow. The ambient temperature at the far end of the air inlet is obtained, and the difference between the average temperature of the fins and the ambient temperature at the far end of the air inlet is taken as the effective heat exchange temperature difference. The current cooling contribution of the external forced air cooling circuit is calculated based on the effective cooling air volume, the effective heat exchange temperature difference, the specific heat capacity of air at constant pressure, and the air density. With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated. At the same time, they are converted into corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands, and the corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands are executed.

2. The fully sealed dual-cooled waterproof pump adaptive control system as described in claim 1, characterized in that, Also includes: The waterproof early warning module is configured to: input the leakage signal of the sealing joint surface, the internal humidity of the motor and the running time into the pre-constructed sealing failure prediction model based on the sensor signal, output the evaluation result characterizing the aging degree of the seal and the leakage risk level, and trigger the corresponding graded early warning signal according to the leakage risk level. The online cleaning module is configured to: calculate a first temperature difference based on the sensor signal, using the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet; calculate a second temperature difference based on the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet; and calculate the current actual heat dissipation efficiency based on the second temperature difference and the current cooling contribution of the external forced air cooling circuit.

3. The fully sealed dual-cooled waterproof pump adaptive control system as described in claim 2, characterized in that, The online cleaning module is also configured to: The current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency corresponding to the clean state under the same working conditions to obtain the heat dissipation efficiency attenuation range. Combining the first temperature difference, the heat dissipation efficiency attenuation range and the cumulative running time, the degree of external pollution of the water pump is determined. When the degree of pollution reaches the preset cleaning trigger threshold, the online cleaning program is started to control the start and stop of the cleaning device, the cleaning water pressure and the cleaning time.

4. The fully sealed dual-cooled waterproof pump adaptive control system as described in claim 3, characterized in that, The target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit include: Obtain the total heat load, the current cooling contribution of the in-shaft cooling circuit, and the current cooling contribution of the external forced air cooling circuit; within a preset in-shaft cooling flow rate adjustment range, discretize into multiple candidate in-shaft cooling contributions according to the first step length; within a preset cooling fan speed adjustment range, discretize into multiple candidate external air cooling contributions according to the second step length; combine the multiple candidate in-shaft cooling contributions and the multiple candidate external air cooling contributions in pairs to form a candidate allocation scheme set.

5. The fully sealed dual-cooled waterproof pump adaptive control system as described in claim 4, characterized in that, The target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit also include: For each candidate allocation scheme in the candidate allocation scheme set, based on the temperature rise response matrix preset in the adjustment module, the predicted temperature rise of the motor stator winding, the predicted temperature rise of the front bearing, and the predicted temperature rise of the rear bearing under the candidate allocation scheme are calculated respectively; the temperature rise response matrix represents the independent influence coefficient and cross influence coefficient of the unit shaft cooling contribution increment and the unit external air cooling contribution increment on the temperature rise of the motor stator winding, the front bearing, and the rear bearing.

6. The fully sealed dual-cooled waterproof pump adaptive control system as described in claim 5, characterized in that, The target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit also include: Remove any candidate allocation scheme whose predicted temperature rise exceeds its corresponding safety threshold from the candidate allocation scheme set, and retain the candidate allocation schemes in which the predicted temperature rise does not exceed the corresponding safety threshold to form a feasible allocation scheme set. For each feasible allocation scheme in the set of feasible allocation schemes, based on the energy consumption function of the in-shaft cooling circuit and the energy consumption function of the external forced air cooling circuit preset in the adjustment module, the estimated energy consumption of the in-shaft cooling circuit and the estimated energy consumption of the external forced air cooling circuit under the corresponding feasible allocation scheme, as well as the estimated total cooling energy consumption of each feasible allocation scheme, are calculated respectively; the energy consumption function of the in-shaft cooling circuit is a function of the contribution of in-shaft cooling, and the energy consumption function of the external forced air cooling circuit is a function of the contribution of external air cooling.

7. The fully sealed dual-cooled waterproof pump adaptive control system as described in claim 6, characterized in that, The target cooling contribution of the internal cooling circuit and the target cooling contribution of the external forced air cooling circuit also include: From the set of feasible allocation schemes, the feasible allocation scheme with the minimum estimated total cooling energy consumption is selected as the optimal allocation scheme. The candidate in-shaft cooling contribution in the optimal allocation scheme is taken as the target cooling contribution of the in-shaft cooling loop, and the candidate external air cooling contribution in the optimal allocation scheme is taken as the target cooling contribution of the external forced air cooling loop. If the set of candidate allocation schemes is empty after elimination, then the boundary processing strategy is executed: the maximum output cooling contribution of the internal cooling circuit and the maximum output cooling contribution of the external forced air cooling circuit are respectively used as the target cooling contribution and output, and at the same time, the insufficient cooling capacity alarm signal is triggered.

8. An adaptive control method for a fully sealed dual-cooled waterproof pump, implemented based on the fully sealed dual-cooled waterproof pump adaptive control system according to any one of claims 1-7, characterized in that, include: Real-time acquisition of sensor signals deployed at key locations on the water pump; Based on the sensor signals, the motor input power is calculated using the motor current and motor voltage, and the total heat load is determined using the motor stator winding temperature, front bearing temperature, rear bearing temperature, and motor input power. Meanwhile, the current cooling contribution of the internal cooling circuit is calculated from the difference between the inlet and outlet temperatures of the internal cooling circuit and the flow rate of the internal cooling circuit. The current cooling contribution of the external forced air cooling circuit is calculated from the difference between the ambient temperature at the far end of the air inlet and the temperature of the heat dissipation fins and the speed of the cooling fan blades. With the goal of minimizing total cooling energy consumption, and under the constraint that the temperature rise of the motor stator winding, front bearing, and rear bearing does not exceed the corresponding safety threshold, the target cooling contribution of the in-shaft cooling circuit and the target cooling contribution of the external forced air cooling circuit are allocated. At the same time, they are converted into corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands, and the corresponding in-shaft cooling flow rate adjustment commands and cooling fan speed adjustment commands are executed.

9. The adaptive control method for a fully sealed dual-cooled waterproof pump as described in claim 8, characterized in that, Also includes: Based on the sensor signals, the leakage signal of the sealing joint surface, the internal humidity of the motor and the running time are input into the pre-built sealing failure prediction model, and the assessment results characterizing the aging degree of the seal and the leakage risk level are output. The corresponding graded early warning signal is triggered according to the leakage risk level. Based on the sensor signal, the first temperature difference is calculated from the difference between the surface temperature of the casing and the ambient temperature at the far end of the air inlet, the second temperature difference is calculated from the difference between the temperature of the heat dissipation fins and the ambient temperature at the far end of the air inlet, and the current actual heat dissipation efficiency is calculated from the second temperature difference and the current cooling contribution of the external forced air cooling circuit. The current actual heat dissipation efficiency is compared with the benchmark heat dissipation efficiency corresponding to the clean state under the same working conditions to obtain the heat dissipation efficiency attenuation range. Combining the first temperature difference, the heat dissipation efficiency attenuation range and the cumulative running time, the degree of external pollution of the water pump is determined. When the degree of pollution reaches the preset cleaning trigger threshold, the online cleaning program is started to control the start and stop of the cleaning device, the cleaning water pressure and the cleaning time.

Images