Intelligent heat supply energy-saving control method based on deep learning

Abstract

The invention discloses an intelligent heat supply energy-saving control method based on deep learning, particularly relates to the technical field of water source heat pump heat supply, and is used for solving the problems of sudden increase of invalid power consumption of a compressor and recessive attenuation of energy efficiency caused by a monitoring blind area of a water body temperature field in the prior art. The method comprises the following steps: acquiring upstream water environment parameters of a heat absorption port in real time, and generating a downstream dynamic temperature distribution state by adopting deep learning enhanced dynamic heat transfer coupling analysis; on the basis of thermal gradient vortex structure recognition and water body heat capacity supplement rate unbalance judgment, a local low-temperature area is accurately positioned; establishing an electric hydraulic frequency domain coupling relation between the heat pump operation frequency and the water body flow velocity pulsation spectrum, and quantifying the hydraulic resonance triggering probability; ineffective power consumption increment is evaluated in combination with the low-temperature region state parameters and the resonance probability; and finally, the driving frequency of the compressor is dynamically adjusted according to the condition that the invalid power consumption increment exceeds the threshold value, and the flow of the circulating pump is synchronously optimized, so that full-chain control from local low-temperature region identification to energy efficiency closed-loop optimization is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water source heat pump heating, and particularly relates to a smart heating energy-saving control method based on deep learning. BACKGROUND

[0002] In a water source heat pump heating system using an open water body (such as a lake or a river), energy transfer is achieved by heat absorption from the water body, which is a core energy-saving means. The existing technology generally relies on a single or sparse temperature monitoring point to feed back the state of the water body, and controls the heat pump unit based on this. This mode assumes that the temperature field of the water body is uniform, and the control strategy is based on a fixed heat conduction model, which maintains system operation by adjusting the power or flow of the compressor.

[0003] Due to the dynamic flow characteristics of the open water body and the continuous heat absorption effect of the heat pump, a significant local low-temperature zone will be formed in the downstream area of the heat absorption port of the heat pump in actual operation. However, the upstream monitoring point relied on by the existing control method cannot sense the state of the low-temperature zone, resulting in that when the system continuously operates at high power, the actual heat absorption temperature is much lower than the feedback value of the control system. This uneven temperature field distribution phenomenon causes a sharp increase in invalid power consumption of the heat pump compressor, and the system energy efficiency is implicitly attenuated, and cannot be identified or corrected by conventional control strategies. SUMMARY

[0004] The present application provides a smart heating energy-saving control method based on deep learning to solve the technical problems in the prior art.

[0005] The technical solution of the present application to solve the above technical problems is as follows:

[0006] A smart heating energy-saving control method based on deep learning, comprising:

[0007] S1, acquiring water body environmental parameters in the upstream area of the heat absorption port of the water source heat pump in real time;

[0008] S2, analyzing the interaction between the water flow path and the heat absorption effect of the heat pump based on the water body environmental parameters through a deep learning enhanced dynamic heat transfer coupling process, and generating a dynamic temperature distribution state in the downstream area of the heat absorption port;

[0009] S3, identifying the thermal gradient vortex structure and calculating the water body heat capacity replenishment rate according to the dynamic temperature distribution state, and determining that there is a local low-temperature zone in the downstream area of the heat absorption port when the vortex center temperature of the thermal gradient vortex structure is lower than a critical value and the water body heat capacity replenishment rate is imbalanced;

[0010] S4, when there is a local low-temperature zone, establishing an electrical and hydraulic frequency domain coupling relationship by acquiring heat pump operation frequency combination parameters and real-time water flow velocity fluctuation spectrum, and evaluating the probability of triggering hydraulic resonance;

[0011] S5, combine the state parameters of the local low temperature area with the hydraulic resonance triggering probability to evaluate the invalid power consumption increment of the heat pump compressor;

[0012] S6, based on the comparison result of the invalid power consumption increment and the preset energy efficiency threshold, dynamically adjust the operation power parameter of the heat pump compressor.

[0013] Further, the water body environmental parameters of the area upstream of the heat absorption port of the water source heat pump are acquired in real time, including:

[0014] The water body temperature monitoring point data is acquired by a temperature sensor array, and the temperature sensor array is arranged on the water flow path of the area upstream of the heat absorption port;

[0015] The water body flow rate monitoring point data is acquired by a flow rate sensor, and the flow rate sensor is arranged at the same point as the temperature sensor;

[0016] The meteorological historical data is acquired by a meteorological monitoring station, and the meteorological historical data includes continuous recorded values of air temperature, wind speed and sunshine intensity in a historical period;

[0017] The water body temperature monitoring point data, the water body flow rate monitoring point data and the meteorological historical data are aligned according to the time stamp to form a synchronous water body environmental parameter set.

[0018] Further, the interaction between the water flow path and the heat pump heat absorption effect is analyzed based on the water body environmental parameters through a deep learning enhanced dynamic heat transfer coupling process to generate a dynamic temperature distribution state of the area downstream of the heat absorption port, including:

[0019] The heat transfer coupling relationship of the water flow path is constructed;

[0020] The spatial distribution characteristics of the water body temperature monitoring point data, the water body flow rate monitoring point data and the air temperature historical data are processed by the trained deep learning network to output a turbulent heat diffusion coefficient correction value;

[0021] The turbulent heat diffusion coefficient correction value is updated to the heat diffusion parameter in the heat transfer coupling relationship;

[0022] Based on the updated heat transfer coupling relationship, the temperature change process of each position point in the area downstream of the heat absorption port is calculated to generate a dynamic temperature distribution state.

[0023] Further, the heat transfer coupling relationship includes an influence factor of the heat pump heat absorption effect on the water body temperature.

[0024] Further, the heat gradient vortex structure is identified and the water body heat capacity replenishment rate is calculated according to the dynamic temperature distribution state, when the vortex center temperature of the heat gradient vortex structure is lower than a critical value and the water body heat capacity replenishment rate is unbalanced, it is determined that there is a local low temperature area in the area downstream of the heat absorption port, including:

[0025] extracting a temperature gradient vector field from the dynamic temperature distribution state;

[0026] identifying a closed ring-shaped isotherm structure based on the temperature gradient vector field, and marking a center point of the closed ring-shaped isotherm structure as a vortex center;

[0027] obtaining a temperature value corresponding to the vortex center as a vortex center temperature;

[0028] calculating a heat exchange amount of the water flowing through the heat pump heat absorption port per unit time as a heat capacity supplement rate reference value;

[0029] real-time monitoring of the heat of the water flowing into the heat absorption port as an actual heat capacity supplement rate;

[0030] when the actual heat capacity supplement rate is less than the heat capacity supplement rate reference value, determining that the water heat capacity supplement rate is imbalanced;

[0031] when the vortex center temperature is lower than a preset critical temperature threshold and the water heat capacity supplement rate is imbalanced, confirming that there is a local low temperature area downstream of the heat absorption port.

[0032] Further, when there is a local low temperature area, an electrical and hydraulic frequency domain coupling relationship is established between the heat pump operating frequency combination parameters and the real-time water flow velocity fluctuation spectrum to evaluate the hydraulic resonance triggering probability, including:

[0033] According to the determination result of the existence of the local low temperature area, the heat pump operating frequency monitoring is started to obtain the heat pump operating frequency combination parameters, including the compressor driving frequency and the circulating pump working frequency;

[0034] Based on the water flow velocity monitoring point data, time-frequency conversion processing is performed to generate a real-time water flow velocity fluctuation spectrum;

[0035] The heat pump operating frequency combination parameters and the real-time water flow velocity fluctuation spectrum are mapped to the same frequency domain coordinate system;

[0036] Detect the overlap degree of the characteristic frequency point of the heat pump operating frequency combination parameters in the frequency domain coordinate system and the peak frequency interval of the real-time water flow velocity fluctuation spectrum;

[0037] According to the overlap degree of the characteristic frequency point and the peak frequency interval, the hydraulic resonance triggering probability is evaluated.

[0038] Further, the heat pump operating frequency combination parameters and the real-time water flow velocity fluctuation spectrum are mapped to the same frequency domain coordinate system, including:

[0039] Discretize the heat pump operating frequency combination parameters at a preset sampling frequency to generate a compressor driving frequency discrete sequence and a circulating pump working frequency discrete sequence;

[0040] Superimpose the compressor driving frequency discrete sequence and the circulating pump working frequency discrete sequence to generate a comprehensive electrical frequency distribution;

[0041] Normalize the frequency domain coordinates of the real-time water body flow velocity fluctuation spectrum, so that the horizontal coordinate scale is aligned with the frequency scale of the comprehensive electrical frequency distribution.

[0042] Further, in combination with the state parameters of the local low temperature area and the hydraulic resonance triggering probability, the invalid power consumption increment of the heat pump compressor is evaluated, including:

[0043] Obtain the vortex center temperature and water thermal capacity replenishment rate imbalance state of the local low temperature area as the state parameters of the local low temperature area;

[0044] According to the difference between the vortex center temperature and the standard working temperature of the heat pump, the temperature compensation power consumption component is calculated;

[0045] Based on the water thermal capacity replenishment rate imbalance state, the heat exchange efficiency loss power consumption component is determined;

[0046] According to the hydraulic resonance triggering probability, the resonance additional power consumption component is calculated;

[0047] Superimpose the temperature compensation power consumption component, the heat exchange efficiency loss power consumption component and the resonance additional power consumption component to generate the invalid power consumption increment of the heat pump compressor.

[0048] Further, according to the hydraulic resonance triggering probability, the resonance additional power consumption component is calculated, including:

[0049] Obtain the mechanical vibration basic power consumption value of the heat pump compressor under the rated working condition;

[0050] Establish a linear mapping relationship between the hydraulic resonance triggering probability and the vibration power consumption amplification coefficient;

[0051] Based on the vibration power consumption amplification coefficient, the mechanical vibration basic power consumption value is proportionally amplified to generate the resonance additional power consumption component.

[0052] Further, based on the comparison result of the invalid power consumption increment and the preset energy efficiency threshold, the operating power parameter of the heat pump compressor is dynamically adjusted, including:

[0053] Input the invalid power consumption increment of the heat pump compressor into the preset energy efficiency threshold comparator;

[0054] When the invalid power consumption increment exceeds the preset energy efficiency threshold, the adjustment amount of the operating power parameter of the heat pump compressor is generated;

[0055] According to the adjustment amount, the driving frequency value of the heat pump compressor is reduced;

[0056] Synchronously update the flow control parameter of the heat pump circulating pump, so that the circulating pump flow matches the updated driving frequency value;

[0057] The updated driving frequency value and the circulating pump flow control parameter are output to a heat pump unit execution unit.

[0058] The beneficial effects of the present application are:

[0059] 1. The open water heat pump system energy efficiency is substantially improved by deep learning driven dynamic temperature field reconstruction and multi-physical field coupling control; a real-time identification system of heat absorption port downstream thermal gradient vortex structure is constructed to accurately capture the formation and evolution characteristics of local low temperature area, and the imbalance state of water heat capacity supplement rate is predicted through dynamic heat transfer coupling analysis, so that the hidden thermodynamic abnormality causing invalid power consumption of compressor is identified from the root cause, and the energy efficiency misjudgment problem caused by monitoring blind area in the prior art is solved.

[0060] 2. The electrical and hydraulic frequency domain coupling relationship is established to map the heat pump operation frequency parameter and the water flow velocity fluctuation spectrum to the same analysis dimension, realizing the quantitative evaluation of hydraulic resonance risk; the invalid power consumption increment model of local low temperature area state parameter and resonance probability is combined to accurately quantify the three power consumption components of temperature compensation, heat exchange loss and resonance addition, providing operable decision basis for energy efficiency optimization; finally, through dynamic power closed loop regulation, the operation state of compressor and circulating pump is synchronously coordinated to eliminate the local low temperature area and prevent hydraulic resonance, forming a whole-chain energy efficiency improvement mechanism from phenomenon identification to control execution, significantly improving the working condition adaptability and energy utilization efficiency of open water heat pump system. BRIEF DESCRIPTION OF DRAWINGS

[0061] Figure 1 The flowchart of the present application based on deep learning intelligent heating energy-saving control method is given.

[0062] Figure 2 The flowchart of determining the existence of local low temperature area in the downstream area of heat absorption port is given. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0064] Embodiment: Figure 1 The present application based on deep learning intelligent heating energy-saving control method is given, which comprises:

[0065] S1, real-time acquisition of water environment parameters of water source heat pump heat absorption port upstream area;

[0066] S2, based on the water environment parameter, the dynamic heat transfer coupling process enhanced by deep learning is used to analyze the interaction between the water flow path and the heat pumping heat absorption effect, and the dynamic temperature distribution state of the downstream area of the heat absorption port is generated;

[0067] S3, according to the dynamic temperature distribution state, the thermal gradient vortex structure is identified and the water heat capacity supplement rate is calculated, when the vortex center temperature of the thermal gradient vortex structure is lower than the critical value and the water heat capacity supplement rate is unbalanced, it is determined that there is a local low temperature area in the downstream area of the heat absorption port;

[0068] S4, when there is a local low temperature area, the electrical and hydraulic frequency domain coupling relationship between the heat pump operation frequency combination parameter and the real-time water flow velocity fluctuation spectrum is obtained, and the water resonance triggering probability is evaluated;

[0069] S5, combined with the state parameters of the local low temperature area and the water resonance triggering probability, the invalid power consumption increment of the heat pump compressor is evaluated;

[0070] S6, based on the comparison result of the invalid power consumption increment and the preset energy efficiency threshold, the operation power parameter of the heat pump compressor is dynamically adjusted.

[0071] S1, the water environment parameters of the upstream area of the heat absorption port of the water source heat pump are obtained in real time, and the specific implementation is as follows:

[0072] The water body temperature monitoring point data is obtained by a temperature sensor array, and the temperature sensor array is arranged on the water flow path of the upstream area of the heat absorption port. The temperature sensor array is composed of a plurality of waterproof platinum resistance temperature probes, each probe is arranged on the river cross section along the water flow direction with a preset interval, the probe installation depth is set to the middle layer position of the water body, and the preset interval is dynamically adjusted according to the river width, for example, when the river width is 5-10 meters, the interval is set to 1 meter, and when the river width exceeds 10 meters, the interval is set to 2 meters. The temperature sensor array is connected to the data collector through the RS-485 bus, the data collection frequency is set to once every 10 seconds, the collection accuracy reaches 0.1 degrees Celsius, the collection range covers 0-40 degrees Celsius, and the collected data is transmitted to the central processing unit in real time.

[0073] The water flow velocity monitoring point data is obtained by a flow velocity sensor, and the flow velocity sensor is arranged at the same point as the temperature sensor. The flow velocity sensor adopts an ultrasonic Doppler flow velocity meter, the probe head and the temperature probe are packaged in the same waterproof shell, the flow velocity measurement range is set to 0.01-3 meters per second, and the measurement accuracy is 0.01 meters per second. The flow velocity sensor and the temperature sensor start data collection through the same trigger signal, so that the temperature monitoring point data and the flow velocity monitoring point data recorded at the same time stamp correspond to the same water position.

[0074] The meteorological history data is obtained by a meteorological monitoring station, which is fixedly installed on the shore open area within 50 meters upstream of the heat absorption port. The meteorological history data includes continuous recording values of air temperature, wind speed and sunshine intensity in the history period, and the history period is defined as 24 hours backward from the current time, and the data sampling interval is 1 minute. The air temperature data is collected by a thermocouple sensor in a radiation-proof ventilation cover, the wind speed data is collected by a three-cup anemometer, and the sunshine intensity data is collected by a silicon photodiode radiometer. All the meteorological sensors are time-synchronized with the central processing unit through the network time protocol, and the time synchronization error is controlled within 500 milliseconds.

[0075] The water body temperature monitoring point data, water body flow rate monitoring point data and meteorological history data are aligned by time stamp to form a synchronous water body environment parameter set. The time stamp alignment process includes the following steps: first, the collection time stamps of the temperature sensor array and the flow rate sensor are calibrated through the global positioning system clock signal, and the time deviation after calibration is less than 500 milliseconds; second, the meteorological history data is linearly interpolated according to the minute-level time stamp, and the interpolation method adopts adjacent time point data weighted calculation according to time proportion to generate second-level meteorological data sequence with the same frequency as the water body temperature monitoring point data and the water body flow rate monitoring point data; finally, the water body temperature monitoring point data, the water body flow rate monitoring point data and the interpolated meteorological data are matched according to the millisecond-level time stamp, and stored as a structured database table. The structured database table includes a time stamp primary key field, a water body temperature value field, a water body flow rate value field, an air temperature value field, a wind speed value field and a sunshine intensity value field, and the multi-source data correlation is established through the time stamp primary key to form a synchronous water body environment parameter set with spatiotemporal consistency.

[0076] S2, based on the water body environment parameter, a dynamic heat transfer coupling process enhanced by deep learning is used to analyze the interaction between the water flow path and the heat pump heat absorption effect, and to generate a dynamic temperature distribution state of the downstream area of the heat absorption port, which is implemented as follows:

[0077] A heat transfer coupling relationship of a water flow path is constructed, and the heat transfer coupling relationship includes an influence factor of heat pump heat absorption on water temperature. The heat transfer coupling relationship is established based on the energy conservation principle of fluid mechanics, and specifically represents a partial differential control equation of heat transfer in the water flow process. The control equation is constructed through the following physical processes: the rate of change of water temperature over time is jointly determined by convective heat transfer caused by water flow, thermal diffusion effect, and heat pump heat absorption. Among them, the convective heat transfer term is calculated by the dot product of the water flow velocity vector and the temperature gradient, the thermal diffusion effect term is described by the product of the thermal diffusion coefficient and the second-order derivative of the temperature space, and the heat pump heat absorption term is expressed by a position-dependent source function. The key parameters of the control equation include: the water density is set to 1000 kg / m3, which is used to calculate the heat capacity per unit volume; the specific heat capacity of water is set to 4180 J / (kg·℃), which is used to associate temperature change and heat exchange; the thermal conductivity coefficient is set to 0.6 W / (m·℃), which represents the heat conduction ability of static water. The influence factor of heat pump heat absorption on water temperature is calculated by dividing the rated heat absorption power of the heat pump by the product of the water flow and the specific heat capacity, for example, when the rated heat absorption power of the heat pump is 500 kW and the water flow is 50 L / s, the influence factor is calculated as 10 ℃ / m, and the influence factor is embedded as a position function in the source term. The boundary conditions are set as follows: the inlet boundary adopts Dirichlet condition (determined by upstream water temperature monitoring data), the water surface boundary adopts Robin condition (heat flux is calculated by air temperature, wind speed and water surface temperature difference), and the riverbed boundary adopts Neumann adiabatic condition (normal temperature gradient is zero).

[0078] The influence factor of heat pump heat absorption on water temperature is calculated by the ratio of the rated heat absorption power of the heat pump to the water flow, for example, when the rated heat absorption power of the heat pump is 500 kW and the water flow is 50 L / s, the influence factor is calculated as 10 ℃ / m. The influence factor is embedded in the control equation as a source term, which is used to quantify the disturbance effect of heat pump operation on the water temperature field.

[0079] The deep learning network trained processes spatial distribution characteristics of water temperature monitoring point data, water flow rate monitoring point data and air temperature historical data, and outputs a turbulent heat diffusion coefficient correction value. The training process of the deep learning network includes the following steps: first, a historical data set is constructed, which includes water temperature monitoring point data, water flow rate monitoring point data and air temperature historical data under different seasonal conditions, and the data time span is not less than 6 months; second, spatial feature extraction is performed on the input data, the upstream region of the heat absorption port is divided into grid units, each grid unit is set to 0.5 meters by 0.5 meters, and the mean value of the temperature monitoring point data, the mean value of the flow rate monitoring point data and the sliding average value of the air temperature historical data in each grid unit are aggregated; then the network output target is defined as the measured value of the turbulent heat diffusion coefficient, which is obtained by tracer diffusion experiment, the experimental method is to inject fluorescent dye tracer in the upstream of the heat absorption port, and to set a sensor array downstream of the heat absorption port to monitor the dye concentration distribution, and to calculate the turbulent heat diffusion coefficient according to the concentration space-time change data; finally, a fully connected neural network is constructed, the number of hidden layers is set to 5 layers, the number of neurons in each layer is set to 128, 256, 256, 128 and 64 respectively, the activation function is ReLU function, the training iteration number is set to 1000 times, and the training is stopped when the mean square error of the validation set is less than 0.01. After the deep learning network trained receives real-time input data, it outputs a turbulent heat diffusion coefficient correction value, which is set to 0.1 to 1.5 square meters per second.

[0080] The turbulent heat diffusion coefficient correction value is updated to the heat diffusion parameter in the heat transfer coupling relationship. The update process is as follows: in the control equation of the heat transfer coupling relationship, the original turbulent heat diffusion coefficient is replaced by the turbulent heat diffusion coefficient correction value output by the deep learning network. The original turbulent heat diffusion coefficient is 0.8 square meters per second, and the heat diffusion parameter is dynamically adjusted according to the water flow state after updating, for example, when the water flow rate monitoring point data shows that the flow rate increases by 30%, the turbulent heat diffusion coefficient correction value increases by 15% to 25%. The update operation is triggered and executed after each new monitoring data is obtained.

[0081] Based on the updated heat transfer coupling relationship, the temperature change process of each position point in the downstream area of the heat absorption port is calculated, and a dynamic temperature distribution state is generated. The calculation process adopts the finite volume method for numerical solution, which specifically includes the following steps: firstly, the downstream area of the heat absorption port is discretized into three-dimensional grid elements, the grid element size is set to 0.5 meters by 0.5 meters by 0.5 meters, and the total number of grids is determined according to the actual water area size, for example, 2400 grid elements correspond to a water area of 20 meters by 10 meters by 3 meters; secondly, the boundary conditions are set, the water temperature monitoring point data and the water flow velocity monitoring point data are used as the input of the water inlet boundary, the air temperature and wind speed data in the historical meteorological data are used to calculate the surface heat flux of the water surface boundary, and the heat flux calculation adopts the temperature difference between air temperature and water surface temperature multiplied by the heat convection coefficient; then, time stepping solution is carried out, the time step is set to 1 second, the updated heat transfer coupling equation is solved in each time step, and the temperature change of each grid element is calculated; finally, the temperature values of all grid elements and their change process with time are output, forming a dynamic temperature distribution state, the dynamic temperature distribution state is stored in a three-dimensional array structure, the first dimension of the array corresponds to the X-axis coordinate, the second dimension corresponds to the Y-axis coordinate, and the third dimension corresponds to the time sequence, each array element stores the temperature value of the space position at the corresponding time. The dynamic temperature distribution state is updated every 10 seconds.

[0082] Figure 2 The flow chart for determining the existence of a local low temperature zone in the downstream area of the heat absorption port is given, S3, identifying the thermal gradient vortex structure and calculating the water heat capacity supplement rate according to the dynamic temperature distribution state, when the vortex center temperature of the thermal gradient vortex structure is lower than the critical value and the water heat capacity supplement rate is unbalanced, it is determined that there is a local low temperature zone in the downstream area of the heat absorption port, and the specific implementation is as follows:

[0083] The temperature gradient vector field is extracted from the dynamic temperature distribution state. The extraction process is based on the spatial differentiation calculation of the dynamic temperature distribution state in the three-dimensional array structure, and the specific operation is as follows: for each spatial grid point in the dynamic temperature distribution state, select its adjacent grid elements in the X-axis direction by 0.5 meters, calculate the temperature difference between the two points and divide by 0.5 meters to obtain the temperature change rate component of the point in the X direction; the temperature change rate components in the Y and Z directions are calculated in the same way; the change rate components in the three directions are combined into a temperature gradient vector, and the temperature gradient vectors of all grid points constitute a temperature gradient vector field. The temperature gradient vector field is stored in the form of a three-dimensional vector matrix, the row and column dimensions of the matrix are consistent with the spatial grid dimensions of the dynamic temperature distribution state, and each matrix element stores a three-dimensional vector containing the component values of the temperature gradient in the X, Y and Z directions, and the units of the components are unified as degrees Celsius per meter.

[0084] The temperature gradient vector field is used to identify the closed ring-shaped isotherm structure, and the center point of the closed ring-shaped isotherm structure is marked as the vortex center. The identification process includes the following steps: first, search for a continuous area with consistent vector rotation direction in the temperature gradient vector field, and the consistency of the vector rotation direction is determined by calculating the cosine value of the angle between adjacent vectors. When the cosine value is greater than 0.98, it is determined that the directions are consistent (corresponding to an angle difference of less than 10 degrees); then extract the boundary isotherm of the continuous area, and the isotherm is formed by connecting grid points with the same temperature value. When the isotherm is closed and the area inside the closed ring is greater than 0.25 square meters, it is determined that there is a closed ring-shaped isotherm structure; finally, the geometric center coordinates of the closed ring-shaped isotherm structure are calculated, and the geometric center is obtained by calculating the arithmetic mean of all vertex coordinates of the closed polygon. For example, a quadrilateral isotherm structure has four vertex coordinates (x1, y1), (x2, y2), (x3, y3), and (x4, y4). The center coordinates are ((x1+x2+x3+x4) / 4, (y1+y2+y3+y4) / 4). The center point coordinates are marked as the vortex center.

[0085] The temperature value corresponding to the vortex center is obtained as the vortex center temperature. The acquisition process is as follows: according to the marked vortex center coordinates, the corresponding grid element in the dynamic temperature distribution state three-dimensional array is located. When the vortex center coordinates are located at the grid node, the temperature value of the node is directly read; when the coordinates are located inside the grid element, the bilinear interpolation method is used to calculate: select the four corner grid nodes of the grid element where the center point is located, calculate the reciprocal of the horizontal distance from the center point to each corner point as the weight coefficient, and weight average the temperature values of the four corner points, for example, the temperatures of the four corner points are T1, T2, T3, and T4, and the distances from the center point to each corner point are d1, d2, d3, and d4. The interpolation temperature is (T1 / d1+T2 / d2+T3 / d3+T4 / d4) / (1 / d1+1 / d2+1 / d3+1 / d4).

[0086] The heat exchange amount of the water flowing through the heat pump heat absorption port per unit time is calculated as the heat capacity supplement rate reference value. The heat exchange amount is calculated by the following steps: first, obtain the water flow data from the water flow rate monitoring point data in step S1, multiply the flow rate value by the river cross-sectional area to obtain the volume flow rate, and the river cross-sectional area is determined according to the preset terrain data, for example, the cross-sectional area of a rectangular river with a width of 5 meters and a water depth of 3 meters is 15 square meters; then obtain the water temperature change value, which is the temperature difference between the upstream and downstream of the heat pump heat absorption port, the upstream water temperature is obtained from the temperature monitoring point data closest to the heat absorption port in S1, and the downstream water temperature is obtained from the real-time monitoring value of the temperature sensor installed at the outlet of the heat pump system; finally, calculate the heat exchange amount, the calculation formula is water density x water specific heat capacity x volume flow rate x water temperature change value, wherein the water density is 1000 kg / m3, the water specific heat capacity is 4180 J / (kg·℃), and the calculation result unit is J / s. The heat exchange amount is the heat capacity supplement rate reference value.

[0087] The water inflow heat around the vortex center is monitored in real time as the actual heat capacity supplement rate. The monitoring method is: taking the vortex center coordinates as the center, setting a circular monitoring area with a radius of 1 meter, obtaining the temperature gradient vector field data and temperature data of all grid cells in the circular area; calculating the net flow of the water flowing into the area according to the temperature gradient vector field, specifically calculating the water flux integral through the circular area boundary: discretizing the circular boundary into 360 boundary segments with an arc length of 1 degree, calculating the normal flow velocity component (dot product of flow velocity vector and boundary normal vector) for each boundary segment, multiplying the boundary segment length to obtain the total net flow; the heat calculation uses water density x specific heat capacity x net flow x area average temperature, and the area average temperature is the arithmetic mean of the temperature values of all grid cells in the circular area. The actual heat capacity supplement rate is updated and calculated every 5 seconds.

[0088] When the actual heat capacity supplement rate is less than the heat capacity supplement rate reference value, it is determined that the water heat capacity supplement rate is imbalanced. The determination process sets a dynamic threshold mechanism: first, multiply the heat capacity supplement rate reference value by 0.8 as the imbalance determination threshold, and the coefficient of 80% is set according to the experience value of the minimum heat exchange efficiency allowed by the heat pump system; then monitor the actual heat capacity supplement rate, and when the actual value is lower than the imbalance determination threshold for 3 consecutive monitoring periods (every period is 5 seconds), trigger the water heat capacity supplement rate imbalance determination. The requirement of 3 periods is to avoid misjudgment caused by instantaneous fluctuations.

[0089] When the vortex center temperature is lower than the preset critical temperature threshold and the water body heat capacity supplement rate is unbalanced, it is confirmed that there is a local low temperature area downstream of the heat absorption port. The preset critical temperature threshold is set according to the heat pump anti-freezing protection requirement, and to avoid equipment damage caused by water body icing, a typical value of 5 degrees Celsius is set; the confirmation condition needs to meet two items: first, the vortex center temperature is lower than the critical temperature threshold in the continuous 6 temperature sampling periods (sampling once every 10 seconds), that is, the low temperature state lasts for 1 minute; second, the water body heat capacity supplement rate unbalance determination remains valid during this duration. When the two conditions are met at the same time, a Boolean logic true value is output, confirming that there is a local low temperature area.

[0090] S4, when there is a local low temperature area, the electrical and hydraulic frequency domain coupling relationship between the heat pump operation frequency combination parameter and the real-time water body flow velocity fluctuation spectrum is obtained, and the hydraulic resonance triggering probability is evaluated, which is implemented as follows:

[0091] According to the determination result of the existence of the local low temperature area output by the S3 step, the heat pump operation frequency monitoring is started to obtain the heat pump operation frequency combination parameter. The heat pump operation frequency combination parameter includes the compressor drive frequency and the circulating pump working frequency, which are collected in real time by the following methods: the compressor drive frequency is read through the communication interface connected to the compressor frequency converter, the communication protocol adopts the industrial standard MODBUS-RTU protocol, and the data sampling interval is set to 100 milliseconds; the circulating pump working frequency is obtained through the pump control system, the sampling frequency is set to 50 Hz, and the collection accuracy is 0.01 Hz. The two frequency data are stored as time series data set according to the collection time stamp, and the collection process is automatically activated when the local low temperature area determination result is true, and the collection duration covers at least two complete heat pump operation periods.

[0092] Based on the water body flow velocity monitoring point data obtained in the S1 step, time-frequency conversion processing is performed to generate a real-time water body flow velocity fluctuation spectrum. The processing process includes the following steps: first, the water body flow velocity monitoring point data is preprocessed, and the 3σ criterion is used to eliminate abnormal data points, that is, the data points exceeding the mean value ± 3 times the standard deviation range in the continuous 3 monitoring periods are considered invalid; then the fast Fourier transform algorithm is used to convert the time domain flow velocity signal into frequency domain spectrum, the transform window length is set to 1024 data points, and the Hanning window is used as the window function type to reduce the frequency spectrum leakage; the frequency range of the real-time water body flow velocity fluctuation spectrum is set to 0 Hz to 25 Hz, covering the main energy distribution area of the water body fluctuation, the frequency resolution is set to 0.05 Hz, and the amplitude value is normalized, and the normalized reference is set to 1.0, which is defined as the amplitude value corresponding to 1 meter per second flow velocity.

[0093] The heat pump operation frequency combination parameter and the real-time water flow fluctuation frequency spectrum are mapped to the same frequency domain coordinate system. The mapping process specifically includes: first, the heat pump operation frequency combination parameter is discretized at a preset sampling frequency of 50 Hz. The sampling frequency is selected based on the Shannon sampling theorem (2 times the highest frequency of water fluctuation, which is 25 Hz). The compressor drive frequency time series is sampled at 50 Hz to generate a compressor drive frequency discrete sequence. Similarly, the circulating pump working frequency discrete sequence is generated. Then, the two discrete sequences are superimposed to generate a comprehensive electrical frequency distribution. The superimposing method is to calculate the comprehensive amplitude value of each frequency point. The comprehensive amplitude value is the square root of the sum of the square of the compressor drive frequency amplitude value and the square of the circulating pump working frequency amplitude value. Finally, the real-time water flow fluctuation frequency spectrum is normalized and converted to a frequency domain coordinate. The original frequency scale of the real-time water flow fluctuation frequency spectrum is resampled to the target scale by linear interpolation method. The frequency range of the target scale is set to 0 Hz to 25 Hz, and the frequency interval is set to 0.05 Hz. The interpolation calculation method is as follows: for the target frequency f0, find the two adjacent frequency points f n and f n+1 in the original spectrum (satisfying f n ≤f0≤f n+1 ), the interpolation amplitude A0=A n +(A n+1 -A n )×(f0-f n ) / (f n+1 -f n ), where A n and A n+1 are the amplitude values of frequency points f n and f n+1 , and n is the number of frequency points.

[0094] The degree of overlap between the characteristic frequency points of the heat pump operation frequency combination parameter in the frequency domain coordinate system and the peak frequency interval of the real-time water flow fluctuation frequency spectrum is detected. The detection process includes: characteristic frequency point extraction, identifying significant peak points with amplitude values exceeding 200% of the average amplitude value of the comprehensive electrical frequency distribution from the comprehensive electrical frequency distribution, recording the center frequency value corresponding to the significant peak points as the characteristic frequency points, and setting the threshold value of 200% based on the experience value of heat pump unit vibration alarm; peak frequency interval extraction, marking all frequency points with amplitude values greater than 80% of the maximum amplitude value of the real-time water flow fluctuation frequency spectrum in the real-time water flow fluctuation frequency spectrum, taking the minimum value as the lower boundary and the maximum value as the upper boundary of the peak frequency interval; and overlap degree calculation, counting the number of characteristic frequency points falling within the peak frequency interval, and calculating the percentage of the number in the total number of characteristic frequency points as the overlap degree. For example, if the total number of characteristic frequency points is 5, of which 3 are within the peak frequency interval, the overlap degree is 60%.

[0095] The hydraulic resonance triggering probability is evaluated according to the overlapping degree of the characteristic frequency point and the peak frequency interval. The evaluation rule adopts a segmented function mapping: when the overlapping degree is less than 30%, the hydraulic resonance triggering probability = overlapping degree x 2; when the overlapping degree is greater than or equal to 30% and less than 70%, the hydraulic resonance triggering probability = 60% + (overlapping degree - 30%) x 0.5; when the overlapping degree is greater than or equal to 70%, the hydraulic resonance triggering probability = 80% + (overlapping degree - 70%) x 2. The basis for designing the segmented function is that the resonance risk increases linearly when the overlapping degree is less than 30%, the risk growth slows down when the overlapping degree is between 30% and 70%, and the risk rises sharply when the overlapping degree is more than 70%. The evaluation result is updated and output every 5 seconds, and a high-risk warning signal is generated when the hydraulic resonance triggering probability exceeds 80%.

[0096] S5, in combination with the state parameters of the local low-temperature area and the hydraulic resonance triggering probability, the invalid power consumption increment of the heat pump compressor is evaluated, which is implemented as follows:

[0097] The vortex center temperature of the local low-temperature area and the water thermal capacity supplement rate imbalance state are obtained as the state parameters of the local low-temperature area. The vortex center temperature is directly obtained from the output result of S3 step, specifically the temperature value corresponding to the vortex center point marked in S3 step, which is updated every 10 seconds; the water thermal capacity supplement rate imbalance state is obtained from the water thermal capacity supplement rate imbalance determination result of S3 step, and the determination result is a Boolean logic value, which outputs true value when the actual thermal capacity supplement rate is less than 80% of the thermal capacity supplement rate reference value for 3 consecutive monitoring periods. The state parameters are stored in a structured data form, including two fields: vortex center temperature numerical field (unit: Celsius) and imbalance state flag field (Boolean value), which are updated every time to overwrite the previous data.

[0098] A temperature compensation power consumption component is calculated based on the difference between the vortex center temperature and the standard operating temperature of the heat pump. The calculation process includes the following steps: first, read the standard operating temperature of the heat pump from the heat pump equipment configuration database, which is determined by the equipment manufacturer according to thermodynamic performance test, for example, the standard operating temperature of a certain type of heat pump under rated operating conditions is 12 degrees Celsius; then calculate the temperature difference ΔT = standard operating temperature of heat pump - vortex center temperature, when ΔT ≤ 0, the temperature compensation power consumption component is 0, when ΔT > 0, the subsequent calculation is performed; the temperature compensation power consumption component Ptemp is calculated by the formula Ptemp = K × ΔT × Q, where K is the temperature compensation coefficient (value range 0.04-0.06 kilowatts per degree Celsius per cubic meter per second), Q is the water flow rate (cubic meters per second) calculated by the water body flow rate monitoring point data in step S1. The temperature compensation coefficient K is determined by experimental calibration, and the calibration method is to control the water temperature to be 1-10 degrees Celsius lower than the standard operating temperature in a closed circulating water system, measure the additional power consumption increment required to maintain the rated output power of the heat pump, establish the ΔT-Ptemp relationship curve for linear regression.

[0099] Based on the imbalance state of the water heat capacity replenishment rate, the heat exchange efficiency loss power consumption component is determined. The determination logic is: when the water heat capacity replenishment rate imbalance state is false, the heat exchange efficiency loss power consumption component Ploss = 0; when the water heat capacity replenishment rate imbalance state is true, first calculate the imbalance degree parameter δ = 1-actual heat capacity replenishment rate / heat capacity replenishment rate reference value, then calculate the efficiency loss coefficient η = 0.2 + 0.8 × δ, and finally calculate Ploss = η × Pnominal. Where Pnominal is the rated power of the heat pump (kilowatts), 0.2 is the fixed heat loss coefficient (reflecting the inherent heat dissipation of the system), and 0.8 is the adjustable efficiency factor (determined by historical operation data analysis and fitting). For example, when δ = 0.5, η = 0.2 + 0.8 × 0.5 = 0.6, and if Pnominal = 1000 kilowatts, then Ploss = 600 kilowatts. This component is refreshed every 5 seconds with the imbalance state update.

[0100] The mechanical vibration basic power consumption value of the heat pump compressor under rated operating conditions is obtained. The acquisition method is: during the heat pump factory test stage, the equipment is installed on a vibration isolation base, and the vibration related power consumption component of the compressor under rated load and no water flow disturbance operating conditions is measured by a high precision power analyzer (measurement error <0.5%). Keep the water temperature constant in the range of standard operating temperature ± 0.5 degrees Celsius, collect 60 minutes of data and take the arithmetic mean as the mechanical vibration basic power consumption value Pvibbase. This value is stored as a device characteristic parameter in the non-volatile memory of the heat pump control system, and the typical value is 3%-5% of the rated power of the heat pump, for example, the unit corresponding to the rated power of 1000 kilowatts is in the range of 30-50 kilowatts.

[0101] A linear mapping relationship between the hydraulic resonance triggering probability and the vibration power consumption amplification coefficient is established. The mapping relationship is defined as: vibration power consumption amplification coefficient β = 1 + α × Pres, where α is the amplification factor (value range 0.7-0.9), and Pres is the hydraulic resonance triggering probability output by S4 step (percentage value divided by 100 to convert to 0-1.0 decimal). The amplification factor α is determined by the vibration table test, and the test method is to simulate 0%-100% resonance probability working conditions on the hydraulic vibration table, measure the ratio of vibration power consumption increment to base power consumption, establish Pres-β relationship curve for linear fitting to obtain the slope α. The linear mapping ensures that when Pres = 0, β = 1.0, and when Pres = 1.0, β = 1.7-1.9.

[0102] The mechanical vibration base power value is proportionally amplified based on the vibration power consumption amplification coefficient to generate a resonance additional power consumption component Presonance, and the generation formula is Presonance = Pvibbase × β, where β is the vibration power consumption amplification coefficient. For example, when Pvibbase = 40 kW and Pres = 0.6, α = 0.8 is taken, then β = 1 + 0.8 × 0.6 = 1.48, and Presonance = 40 × 1.48 = 59.2 kW. This component is refreshed synchronously every 5 seconds with the probability update of S4 step, and when Pres < 0.05, it is considered to have no resonance risk, and Presonance = 0 is directly set.

[0103] The temperature compensation power consumption component, the heat exchange efficiency loss power consumption component, and the resonance additional power consumption component are superimposed to generate the invalid power consumption increment Pinvalid of the heat pump compressor. The superimposition formula is Pinvalid = Ptemp + Ploss + Presonance. The superimposition process adopts a timestamp alignment mechanism: set the valid period identifier for each component (Ptemp valid period 10 seconds, Ploss valid period 5 seconds, Presonance valid period 5 seconds), and only use the latest data within the valid period when performing superimposition calculation; if the data of a component is overdue and not updated, the previous valid value is used and marked as estimated state. The calculation result Pinvalid is output as an instantaneous value (unit: kW), and time series data is recorded at the same time to generate an invalid power consumption trend analysis report. The refresh frequency of the invalid power consumption increment is synchronized with the fastest updating component, that is, a new value is output every 5 seconds.

[0104] S6, based on the comparison result of the invalid power consumption increment and the preset energy efficiency threshold, dynamically adjusting the running power parameter of the heat pump compressor, the specific implementation is as follows:

[0105] The invalid power consumption increment of the heat pump compressor is input into a preset energy efficiency threshold comparator. The preset energy efficiency threshold is set by first reading the theoretical minimum power consumption value of the heat pump unit under the rated operating condition from the heat pump equipment configuration database, which is determined by the equipment manufacturer according to thermodynamic performance testing and stored in the control system; then calling the power consumption record data in the heat pump operation history database, calculating the standard deviation of the operating power consumption in the last 30 days, and taking the theoretical minimum power consumption value plus 3 times the standard deviation as the preset energy efficiency threshold baseline value; finally, dynamically adjusting according to the seasonal operating condition factor, setting the summer operating condition factor to 1.0 and the winter operating condition factor to 1.2, for example, the preset energy efficiency threshold is equal to the baseline value multiplied by 1.2 when operating in winter. The comparator uses a window comparison algorithm and sets a continuous determination mechanism: when the invalid power consumption increment of the compressor exceeds the threshold value for 3 consecutive sampling periods, the determination is triggered, and the sampling period is set to 5 seconds, which is synchronized with the invalid power consumption increment update frequency.

[0106] When the invalid power consumption increment exceeds the preset energy efficiency threshold, the adjustment amount of the heat pump compressor operating power parameter is generated. The adjustment amount generation process includes: first, calculating the proportion parameter of the invalid power consumption increment exceeding the threshold, which is equal to the difference between the invalid power consumption increment and the preset energy efficiency threshold divided by the preset energy efficiency threshold; then determining the frequency adjustment amount according to the proportion parameter, the adjustment rule uses a piecewise function: when the proportion parameter is less than or equal to 0.2, the adjustment amount is equal to the proportion parameter multiplied by the maximum allowed adjustment amount multiplied by 0.5; when the proportion parameter is greater than 0.2 and less than or equal to 0.5, the adjustment amount is equal to the proportion parameter multiplied by the maximum allowed adjustment amount multiplied by 0.8; when the proportion parameter is greater than 0.5, the adjustment amount is equal to the proportion parameter multiplied by the maximum allowed adjustment amount, where the maximum allowed adjustment amount is set to 20% of the current drive frequency value. The adjustment amount generation process sets a gradient protection mechanism: the single adjustment amount does not exceed 5% of the current drive frequency value, for example, the single maximum adjustment amount is 2.5 Hz when the current drive frequency is 50 Hz. At the same time, set the lower limit of the adjustment amount, when the calculated adjustment amount is less than 0.5 Hz, execute 0.5 Hz.

[0107] According to the adjustment amount, the drive frequency value of the heat pump compressor is reduced. The reduction operation adopts a ramp control strategy: set the frequency change rate not to exceed 1 Hz per second, set the target drive frequency value to the current drive frequency value minus the adjustment amount; when the difference between the target frequency and the current frequency is greater than the frequency change rate, perform the reduction operation in steps, each step reduction value is equal to the frequency change rate, and the step length time is 1 second; when the difference is less than the frequency change rate, directly set it to the target frequency value. The drive frequency value adjustment range is strictly limited within the safe working frequency range specified in the manufacturer's technical manual, for example, a certain type of compressor allows 30-60 Hz range. When the target frequency exceeds the safe range, automatically correct it to the nearest boundary value and generate an out-of-limit alarm record.

[0108] Synchronously update the flow control parameter of the heat pump circulating pump to match the updated drive frequency value. The flow control parameter update is based on the pump similarity law: the ratio of the new flow set value to the adjusted flow value is equal to the ratio of the updated drive frequency value to the adjusted drive frequency value. The flow control parameter is converted to the frequency command of the circulating pump frequency converter, and the conversion relationship is that the circulating pump operating frequency is equal to the flow-frequency conversion coefficient multiplied by the new flow set value, where the flow-frequency conversion coefficient ranges from 0.4 to 0.6 Hz per cubic meter per hour. The conversion coefficient is obtained by pump performance curve calibration: under laboratory conditions, measure the frequency-flow correspondence curve of the circulating pump in the 10% to 100% flow range, and take the average value of the slope of the curve as the conversion coefficient. The update operation is synchronized with the compressor drive frequency adjustment, with a time deviation of not more than 100 milliseconds, ensuring that the hydraulic system responds cooperatively.

[0109] The updated drive frequency value and the circulating pump flow control parameter are output to the heat pump unit execution unit. The output process uses an industrial standard communication protocol: the compressor drive frequency value is written to the compressor frequency converter's holding register address 40001 through the MODBUS-TCP protocol, and the data format is 32-bit floating point number; the circulating pump flow control parameter is converted to a 4-20 milliamp analog signal output to the circulating pump controller, and the conversion method is: the output current value is equal to 4 plus 16 times the circulating pump operating frequency minus the difference between the circulating pump allowed minimum operating frequency, divided by the difference between the circulating pump allowed maximum operating frequency and the minimum operating frequency, where the circulating pump allowed minimum and maximum operating frequencies are, for example, 20 Hz and 50 Hz, respectively. The execution unit completes the device adjustment within 200 milliseconds after receiving the parameters, and the adjustment result is fed back to the control system through the status register. Set up an execution state monitoring mechanism: when the device response times out for 300 milliseconds or returns an error code, automatically resend the control parameters and trigger the device abnormal alarm.

[0110] The scheme of the embodiment dynamically correlates the water thermodynamic properties (vortex center temperature, heat capacity supplement rate) with the equipment electrical parameters (compressor frequency, resonance probability), breaking through the single dimension limitation of only monitoring water temperature or flow in traditional heat pump control. The precise spatial positioning of the local low temperature area is realized through the vortex structure identification of S3 step; the frequency domain coupling analysis of S4 step maps the water pulsation spectrum and the electrical frequency combined parameters to the same coordinate system, revealing the implicit resonance risk. The three power consumption components of temperature compensation, efficiency loss and resonance addition constructed in S5 step are superimposed, solving the problem of accurate decomposition of power consumption increment under complex working conditions. The temperature compensation component is calculated by the dynamic deviation of the vortex center temperature and the standard value, the efficiency loss component is nonlinearly related to the imbalance degree of the heat capacity supplement rate, and the resonance additional component is generated based on the probability driven amplification mechanism. The quantitative conversion of local low temperature area state parameters to invalid power consumption is realized, which provides quantifiable decision basis for energy efficiency optimization. The fluid state diagnosis results are directly converted into the adaptive mechanism of equipment control parameters, overcoming the defects of response lag and ignoring the water characteristics of traditional control strategy.

[0111] The calculations involved in the embodiment are all de-dimensioned numerical calculations, and the preset parameters and threshold values in the calculations are set by those skilled in the art according to the actual situation.

[0112] It should be noted that the present application can be deployed in the device itself to realize embedded application, or run on PC or other terminal with user interface, so as to meet various hardware environments and use requirements.

[0113] The above-described embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented by software, the above-described embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center through wireless or wired transmission. The wired transmission includes optical fiber, twisted pair, coaxial cable, etc. The wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state disk.

[0114] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and module can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0115] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the above-described device embodiments are only schematic, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0116] The modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical modules, which can be located in one place or distributed on a plurality of network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments.

[0117] In addition, each functional module in the various embodiments of the present application can be integrated in one processing module, or each module can exist physically independently, or two or more modules can be integrated in one module.

[0118] If the functions are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0119] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0120] Finally: the above is only the preferred embodiment of the present application, and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A deep learning-based intelligent heating energy-saving control method, characterized in that, Comprise: S1, real-time acquisition of water source heat pump suction port upstream area of water body environmental parameters; S2, based on the water body environmental parameters through the deep learning enhanced dynamic heat transfer coupling process analysis of water flow path and heat pump heat absorption interaction, generate the dynamic temperature distribution state of suction port downstream area; S3, according to the dynamic temperature distribution state identification heat gradient vortex structure and calculate the water heat capacity supplement rate, when the vortex center temperature of heat gradient vortex structure is lower than the critical value and water heat capacity supplement rate imbalance, judge the suction port downstream area exists local low temperature area; S4, when there is local low temperature area, get the heat pump running frequency combination parameter and real-time water flow velocity fluctuation spectrum to establish the electrical and hydraulic frequency domain coupling relationship, evaluate the probability of hydraulic resonance trigger; S5, combined with the state parameters of local low temperature area and hydraulic resonance trigger probability, evaluate the invalid power consumption increment of heat pump compressor; S6, based on the comparison result of invalid power consumption increment and preset energy efficiency threshold, dynamically adjust the running power parameter of heat pump compressor. 2.The deep learning-based intelligent heating energy-saving control method according to claim 1, characterized in that, Real-time acquisition of water source heat pump suction port upstream area of water body environmental parameters, including: Obtain water temperature monitoring point data through temperature sensor array, temperature sensor array is laid on the water flow path of suction port upstream area; Obtain water flow velocity monitoring point data through flow velocity sensor, flow velocity sensor is laid with temperature sensor at the same point; Obtain meteorological historical data through meteorological monitoring station, meteorological historical data includes continuous record value of air temperature, wind speed and sunshine intensity in historical period; Align water temperature monitoring point data, water flow velocity monitoring point data and meteorological historical data by time stamp to form synchronous water body environmental parameter set. 3.The deep learning-based intelligent heating energy-saving control method according to claim 2, characterized in that, Based on the water body environmental parameters through the deep learning enhanced dynamic heat transfer coupling process analysis of water flow path and heat pump heat absorption interaction, generate the dynamic temperature distribution state of suction port downstream area, including: Build the heat transfer coupling relationship of water flow path; Process the spatial distribution characteristics of water temperature monitoring point data, water flow velocity monitoring point data and air temperature historical data through the trained deep learning network, output the turbulence heat diffusion coefficient correction value; Update the turbulence heat diffusion coefficient correction value to the heat diffusion parameter in the heat transfer coupling relationship; Based on the updated heat transfer coupling relationship, calculate the temperature change process of each position point in the suction port downstream area, generate the dynamic temperature distribution state.

4. The deep learning-based intelligent heating energy-saving control method according to claim 3, characterized in that, The heat transfer coupling relationship contains the influence factor of heat pump heat absorption on water temperature.

5. The deep learning-based intelligent heating energy-saving control method according to claim 3, characterized in that, According to the dynamic temperature distribution state, identify the heat gradient vortex structure and calculate the water heat capacity supplement rate, when the vortex center temperature of heat gradient vortex structure is lower than the critical value and water heat capacity supplement rate imbalance, judge the suction port downstream area exists local low temperature area, including: Extract temperature gradient vector field from dynamic temperature distribution state; Identify the closed loop isotherm structure based on temperature gradient vector field, mark the center point of closed loop isotherm structure as vortex center; Get the temperature value corresponding to vortex center as vortex center temperature; Calculate the heat exchange amount of water flowing through heat pump suction port per unit time as heat capacity supplement rate reference value; Real-time monitoring of vortex center surrounding water inflow heat as actual heat capacity supplement rate; When the actual heat capacity supplement rate is less than the heat capacity supplement rate reference value, it is determined that the water body heat capacity supplement rate is imbalanced; When the vortex center temperature is lower than the preset critical temperature threshold and the water body heat capacity supplement rate is imbalanced, it is determined that there is a local low temperature area downstream of the heat absorption port.

6. The deep learning-based intelligent heating energy-saving control method according to claim 5, characterized in that, When there is a local low temperature area, an electrical and hydraulic frequency domain coupling relationship between the heat pump operation frequency combination parameter and the real-time water body flow velocity fluctuation spectrum is obtained, and the hydraulic resonance triggering probability is evaluated, including: According to the determination result of the existence of the local low temperature area, the heat pump operation frequency monitoring is started to obtain the heat pump operation frequency combination parameter, including the compressor driving frequency and the circulating pump working frequency; Based on the time-frequency conversion processing of the water body flow velocity monitoring point data, the real-time water body flow velocity fluctuation spectrum is generated; The heat pump operation frequency combination parameter and the real-time water body flow velocity fluctuation spectrum are mapped to the same frequency domain coordinate system; The overlap degree of the characteristic frequency point of the heat pump operation frequency combination parameter in the frequency domain coordinate system and the peak frequency interval of the real-time water body flow velocity fluctuation spectrum is detected; The hydraulic resonance triggering probability is evaluated according to the overlap degree of the characteristic frequency point and the peak frequency interval.

7. The deep learning-based intelligent heating energy-saving control method according to claim 6, characterized in that, The heat pump operation frequency combination parameter and the real-time water body flow velocity fluctuation spectrum are mapped to the same frequency domain coordinate system, including: The heat pump operation frequency combination parameter is discretized at a preset sampling frequency to generate a compressor driving frequency discrete sequence and a circulating pump working frequency discrete sequence; The compressor driving frequency discrete sequence and the circulating pump working frequency discrete sequence are superimposed to generate a comprehensive electrical frequency distribution; The real-time water body flow velocity fluctuation spectrum is normalized and converted to a frequency domain coordinate, so that the horizontal coordinate scale is aligned with the frequency scale of the comprehensive electrical frequency distribution. 8.The deep learning based intelligent heating energy-saving control method of claim 6, wherein, Combined with the state parameters of the local low temperature area and the hydraulic resonance triggering probability, the invalid power consumption increment of the heat pump compressor is evaluated, including: The vortex center temperature of the local low temperature area and the water body heat capacity supplement rate imbalance state are obtained as the state parameters of the local low temperature area; According to the difference between the vortex center temperature and the standard working temperature of the heat pump, the temperature compensation power consumption component is calculated; Based on the water body heat capacity supplement rate imbalance state, the heat exchange efficiency loss power consumption component is determined; The resonance additional power consumption component is calculated according to the hydraulic resonance triggering probability; The temperature compensation power consumption component, the heat exchange efficiency loss power consumption component and the resonance additional power consumption component are superimposed to generate the invalid power consumption increment of the heat pump compressor. 9.The deep learning based intelligent heating energy-saving control method of claim 8, wherein, The resonance additional power consumption component is calculated according to the hydraulic resonance triggering probability, including: The mechanical vibration basic power consumption value of the heat pump compressor under the rated working condition is obtained; A linear mapping relationship between the hydraulic resonance triggering probability and the vibration power consumption amplification coefficient is established; Based on the vibration power consumption amplification coefficient, the mechanical vibration basic power consumption value is proportionally amplified to generate the resonance additional power consumption component. 10.The deep learning based intelligent heating energy-saving control method of claim 8, wherein, Based on the comparison result of the invalid power consumption increment and the preset energy efficiency threshold, the operation power parameter of the heat pump compressor is dynamically adjusted, including: The invalid power consumption increment of the heat pump compressor is input into the preset energy efficiency threshold comparator; When the invalid power consumption increment exceeds the preset energy efficiency threshold, the adjustment amount of the heat pump compressor operation power parameter is generated; According to the adjustment amount, the driving frequency value of the heat pump compressor is reduced; The flow control parameter of the heat pump circulating pump is updated synchronously, so that the circulating pump flow matches the updated driving frequency value; The updated drive frequency value and the circulating pump flow control parameter are output to a heat pump unit execution unit. The updated drive frequency value and the circulating pump flow control parameter are output to a heat pump unit execution unit.

Images