An adaptive control method for an intelligent coupled heat exchanger unit

Abstract

The application relates to the technical field of heat exchanger unit control, and discloses an adaptive control method for an intelligent coupled heat exchanger unit, which comprises the following steps: acquiring a heat pump instantaneous COP value, an evaporator side primary return water temperature drop, an outdoor temperature prediction curve and a node measured temperature; calculating a deviation based on the instantaneous COP value and the primary return water temperature drop, determining an electric regulating valve opening degree adjustment amount through fuzzy reasoning; determining a future heat load prediction value, combining a secondary water supply temperature deviation to determine a final heat pump compressor operating frequency; triggering an anti-freezing response operation based on the evaporator outlet temperature, controlling the electric regulating valve and a bypass short circuit valve; and dividing a working condition interval according to the secondary return water temperature and calling a matched gain coefficient to adjust the frequency. The application can reduce pipe network hydraulic disturbance, improve the secondary side water supply temperature, dynamically optimize the shunt ratio to maintain the unit energy efficiency, shorten the load mutation response time, prevent the evaporator from icing under extreme working conditions and inhibit temperature fluctuation.

Description

Technical Field

[0001] This invention relates to the field of heat exchanger control technology, specifically to an adaptive control method for an intelligent coupled heat exchanger. Background Technology

[0002] Urban centralized heating systems are a crucial infrastructure in modern cities. Their core operation involves transferring high-temperature heat from the primary heating network to the secondary network via heat exchange stations to serve end users. In this heat transfer process, the heat exchange unit, as the core equipment, typically consists of key components such as plate heat exchangers, circulating water pumps, and control systems. Furthermore, with the increasing demand for energy conservation and emission reduction, deep heat extraction and large temperature difference heating are gradually becoming the main development directions in this field, aiming to further reduce the return water temperature of the primary network and recover low- and medium-grade waste heat resources.

[0003] Currently, existing technical solutions in this field mainly fall into two mainstream categories. The first is traditional plate heat exchanger technology, where primary water supply is returned to the heat source plant at a medium temperature after releasing heat through the plate heat exchanger. The second is integrated energy utilization technology combining plate heat exchangers and water source heat pumps. This type of solution typically connects the entire primary return water flow in series to the evaporator of the water source heat pump system for cooling, and connects the condenser to the primary side supply water pipeline to extract heat. Regarding system control strategies, existing units mainly employ conventional reactive control logic, adjusting the heat output through PID algorithms based on real-time detected temperature or pressure deviations. Furthermore, the water flow split ratio between the heat pump and the main pipeline is often set to a fixed empirical value.

[0004] However, in terms of hydraulic and thermal structure design, the existing technology, which uses a single-pass return water series connection to the evaporator, significantly increases the additional resistance of the pipe network system and causes considerable disturbance to the water circulation conditions of the existing pipe network. Furthermore, the condenser's method of extracting heat from the primary side makes it difficult for the extracted waste heat to directly serve secondary users, resulting in a less than ideal increase in secondary water supply temperature. Secondly, at the control method level, the existing fixed split ratio setting is difficult to adapt to dynamic changes in outdoor climate conditions and heating load, easily causing the heat pump unit to deviate from its optimal energy efficiency range. Moreover, conventional reactive control logic exhibits lag in response to sudden load changes, easily causing fluctuations in user-end heating temperature. In addition, the existing anti-freeze protection strategy still has shortcomings in its dynamic adaptive adjustment mechanism under extreme conditions. Therefore, an adaptive control method for intelligent coupled heat exchanger units is urgently needed to solve the above problems. Summary of the Invention

[0005] To address the problems in related technologies, this invention provides an adaptive control method for intelligent coupled heat exchanger units to overcome the aforementioned technical problems in existing related technologies.

[0006] To solve the aforementioned technical problem, the present invention is achieved through the following technical solution: In a first aspect, embodiments of the present invention provide an adaptive control method for an intelligent coupled heat exchanger unit, specifically including: Obtain the instantaneous COP value of the heat pump of the heat exchange unit, the temperature drop of the primary return water on the evaporator side, the outdoor temperature prediction curve, the actual temperature of the secondary water supply, the evaporator outlet temperature and the secondary return water temperature; Based on the instantaneous COP value of the heat pump and the temperature drop of the primary return water on the evaporator side, the heat pump energy efficiency deviation and the evaporator temperature drop deviation are calculated. The opening adjustment amount of the electric regulating valve is determined by fuzzy reasoning. The opening is adjusted according to the opening adjustment amount, and the system returns to continue to obtain the instantaneous COP value of the heat pump and the temperature drop of the primary return water on the evaporator side to form a closed-loop control. The system acquires outdoor temperature forecast, solar radiation forecast, and wind speed forecast, determines future heat load forecast, and performs feedback correction based on the temperature deviation between the actual secondary water supply temperature and the actual temperature setpoint to determine the final operating frequency of the heat pump compressor. Based on the evaporator outlet temperature, a corresponding antifreeze response operation is triggered to control the electric regulating valve and bypass short-circuit valve on the first bypass branch. Based on the secondary return water temperature, the operating range is divided, and the operating frequency of the heat pump compressor is adjusted by calling the matching proportional gain coefficient, integral gain coefficient, and derivative gain coefficient.

[0007] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, the method involves calculating the heat pump efficiency deviation and the evaporator temperature drop deviation based on the instantaneous COP value of the heat pump and the primary return water temperature drop on the evaporator side, and determining the adjustment amount of the electric regulating valve opening through fuzzy reasoning, including: The difference between the instantaneous COP value of the heat pump and the preset heat pump energy efficiency target value is used to obtain the heat pump energy efficiency deviation. The temperature drop deviation of the evaporator is obtained by subtracting the temperature drop of the primary return water on the evaporator side from the preset target value of the temperature drop of the primary return water on the evaporator side. Calculate the quantization factors of the heat pump energy efficiency deviation and the evaporator temperature drop deviation, as well as the proportional factor of the electric regulating valve opening adjustment amount. Map the physical domains of the heat pump energy efficiency deviation, the evaporator temperature drop deviation and the electric regulating valve opening adjustment amount to a fuzzy domain interval and divide them into multiple fuzzy subsets. Using the heat pump energy efficiency deviation and evaporator temperature drop deviation as input variables for fuzzy inference, the precise fuzzy value of the electric regulating valve opening adjustment amount is output according to the fuzzy inference rule base. The precise fuzzy value is multiplied by the proportional factor to calculate the electric regulating valve opening adjustment amount.

[0008] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, the precise fuzzy value of the electric regulating valve opening adjustment amount is output according to the fuzzy inference rule base, including: A fuzzy inference rule base containing multiple control rules is established. The evaporator temperature drop deviation, which reflects the hydraulic and thermodynamic balance of the system, is used as the dominant variable to determine the direction of the opening adjustment. The heat pump energy efficiency deviation is used as the constraint variable to perform asymmetric dynamic correction of the adjustment range. When the heat pump energy efficiency deviation belongs to the negative subset, if the evaporator temperature drop deviation is positive, a command is output to accelerate the opening of the electric regulating valve; if the evaporator temperature drop deviation is negative, the closing rate of the electric regulating valve is forcibly reduced. When the heat pump energy efficiency deviation belongs to a positive subset, a convergence strategy of rapid shutdown and buffered startup is adopted. The centroid method is used to defuzzify the output fuzzy set obtained from fuzzy inference, and the precise fuzzy value of the opening adjustment amount of the electric regulating valve is calculated.

[0009] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit of the present invention, determining the predicted future heat load based on the outdoor temperature prediction curve includes: Historical outdoor temperature sequences, historical secondary water supply temperature sequences, and historical measured heating loads were collected as feature variables and packaged into a time series input tensor using the sliding window method. The time series input tensor is fed into a long short-term memory network for processing. The network is then trained offline using a comprehensive loss function that incorporates physical constraints related to building thermal inertia, and the predicted future heat load is output. The comprehensive loss function includes a mean square error term and a physical constraint penalty term constructed based on the maximum allowable rate of change of load between adjacent time steps.

[0010] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, determining the operating frequency of the heat pump compressor and the opening degree of the diverter valve based on the predicted future heat load includes: A cost function is constructed with the goal of minimizing the total input power of the system under the predicted heat load. Under the constraints of the set feasible region, the target heat pump load sharing ratio is extracted by optimization. The preset heating capacity of the heat pump at future time is determined based on the product of the target heat pump load sharing ratio and the predicted value of the future heat load. Substitute the preset heating capacity of the heat pump at future time into the fitting equation of the compressor's frequency conversion characteristics to determine the operating frequency of the heat pump compressor. The instantaneous COP value of the heat pump collected in real time is processed by a first-order low-pass filter algorithm to calculate the effective heating performance coefficient of the heat pump at the current moment. The target diversion flow rate is determined by combining the preset heating capacity of the heat pump at the future moment, the effective heating performance coefficient of the heat pump at the current moment, the specific heat capacity of the circulating water at constant pressure, the density of the circulating water, and the preset target value of the primary return water temperature drop on the evaporator side. The opening degree of the diversion valve on the first bypass branch is determined by using the inverse mapping function between the target diversion flow rate and the equal percentage flow characteristic of the electric regulating valve, combined with the piecewise calculation function of the boundary constraint logic.

[0011] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, the determination of the final operating frequency of the heat pump compressor includes: Extract the temperature deviation of the previous time step and the temperature deviation of the previous two time steps, and combine the temperature deviation, proportional gain coefficient, integral gain coefficient and derivative gain coefficient to calculate the frequency feedback correction using an incremental PID algorithm. The frequency feedback correction is limited based on the set temperature control dead zone threshold and the maximum allowable frequency adjustment step size in a single operation. The final operating frequency of the heat pump compressor is determined by adding the operating frequency of the heat pump compressor to the frequency feedback correction amount after amplitude limiting.

[0012] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, the step of triggering a corresponding antifreeze response operation based on the evaporator outlet temperature includes: Collect the evaporator outlet temperature and set warning level threshold, danger level threshold and recovery level threshold; When the evaporator outlet temperature exceeds the warning level threshold, normal operation is maintained. When the evaporator outlet temperature is greater than the danger level threshold and less than or equal to the warning level threshold, a warning state response operation is triggered, which increases the opening of the electric regulating valve on the first bypass branch and decreases the operating frequency of the heat pump compressor. When the evaporator outlet temperature is less than or equal to the danger level threshold, a danger state response operation is forcibly triggered, the heat pump compressor and the electric regulating valve are shut down, the bypass short-circuit valve is opened to establish internal water self-circulation, and the danger state response operation is locked while the evaporator outlet temperature is less than the recovery level threshold. When the evaporator outlet temperature is greater than or equal to the recovery stage threshold and continues for a set time threshold, the state lock is released and the step-by-step recovery program is started. The bypass short-circuit valve is closed and the opening of the electric regulating valve is increased step by step. After the water temperature stabilizes, the heat pump compressor is restarted and the operating frequency is restored step by step. Closed-loop monitoring is performed during the execution of the step-by-step recovery procedure. If the evaporator outlet temperature drops to the warning level threshold or the danger level threshold, the step-by-step recovery procedure is immediately interrupted and the corresponding antifreeze response operation is rolled back.

[0013] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, the method involves dividing the operating range based on the secondary return water temperature and adjusting the operating frequency of the heat pump compressor by calling the matched proportional gain coefficient, integral gain coefficient, and derivative gain coefficient, including: Based on the set low load zone threshold and high load zone threshold, the operating conditions are divided into low load zone, normal load zone and high load zone according to the secondary return water temperature, and independent proportional gain coefficient, integral gain coefficient and differential gain coefficient are configured for low load zone, normal load zone and high load zone respectively. Determine the operating range to which the secondary return water temperature belongs, and extract the corresponding proportional gain coefficient, integral gain coefficient, and differential gain coefficient. The setpoint of the secondary water supply temperature is used as the control target, and the actual temperature of the secondary water supply is used as the controlled variable. An incremental PID algorithm is used to calculate the frequency adjustment of the heat pump compressor.

[0014] As a preferred embodiment of the adaptive control method for the intelligent coupled heat exchanger unit described in this invention, the process of extracting the corresponding proportional gain coefficient, integral gain coefficient, and differential gain coefficient further includes: when the secondary return water temperature crosses the boundary point of the operating condition range, the proportional gain coefficient, integral gain coefficient, and differential gain coefficient perform linear gradual change within a set gradual change time based on the difference between the current coefficient value and the target coefficient value.

[0015] Secondly, embodiments of the present invention provide an intelligent coupled heat exchange unit, comprising: a plate heat exchanger, a water source heat pump unit, a first bypass branch, a second bypass branch, an intelligent coupled controller, a temperature sensor, a pressure sensor, a flow meter, a variable frequency circulating pump, and an electric regulating valve. The water source heat pump unit includes an evaporator, a compressor, a condenser, and an expansion valve; The evaporator is connected in parallel to the primary return water pipeline via a first bypass branch, and the condenser is connected in parallel to the secondary return water pipeline via a second bypass branch. Both the first and second bypass branches adopt a branch pipe structure that is inclined to the main pipe. The intelligent coupling controller is connected to the temperature sensor, pressure sensor, flow meter, variable frequency circulating pump, electric regulating valve, and compressor frequency converter of the water source heat pump unit.

[0016] The present invention has the following beneficial effects: 1. This invention employs a structure in which the first and second bypass branches are connected in parallel at an oblique angle to the primary return water main and the secondary return water main, respectively. This diverts a portion of the primary return water into the evaporator for cooling, and a portion of the secondary return water into the condenser for heat absorption and heating. The branch diversion structure reduces the pressure drop in the main flow path of the heat exchanger unit, minimizes the impact on the operating resistance of the existing network's variable frequency circulating pump, and achieves hydraulic decoupling between the plate heat exchanger and the water source heat pump unit. Simultaneously, the condenser releases the heat extracted from the primary return water to the secondary return water, increasing the overall secondary supply water temperature and overcoming the drawbacks of increased network resistance and limited secondary-side heating temperature caused by the full-flow series connection.

[0017] 2. In this invention, the intelligent coupling controller collects the heat pump energy efficiency deviation and evaporator temperature drop deviation and processes them into the fuzzy inference rule base. Based on the output results, it adjusts the opening of the electric regulating valve in real time to control the water flow into the first bypass branch, so that the water source heat pump unit can maintain the target heating performance coefficient under different operating conditions, thus solving the problem that the fixed split ratio setting cannot adapt to the dynamic changes of outdoor load.

[0018] 3. In this invention, the intelligent coupling controller calculates the predicted future heat load by receiving the outdoor temperature prediction curve, sets the compressor's operating frequency and the opening of the electric regulating valve in advance, and combines the incremental PID algorithm based on the actual secondary water supply temperature deviation to output feedback correction to control the compressor's final operating frequency. This allows the unit to complete the pre-setting of equipment parameters before heat load changes occur, shortening the system's control response time to sudden load changes.

[0019] 4. In this invention, the intelligent coupling controller executes corresponding anti-freeze response operations based on the monitored value of the evaporator outlet temperature. By controlling the closure of the electric regulating valve and the opening of the bypass short-circuit valve on the first bypass branch, it maintains the self-circulation flow of the water source heat pump unit on the condenser side, preventing the low-temperature circulating water from freezing inside the evaporator and ensuring uninterrupted secondary heating of the heat exchange unit under extreme operating conditions. Furthermore, the intelligent coupling controller divides different operating ranges according to the secondary return water temperature and calls the corresponding parameter groups. When switching between operating ranges, it performs a linear gradual change in control parameters, reducing fluctuations in the secondary supply water temperature during the control logic transition process.

[0020] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, the drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 The schematic diagram of the intelligent coupled heat exchanger system provided by the present invention.

[0023] Figure 2 This is a schematic diagram of the hydraulic structure of the branch flow path provided by the present invention.

[0024] Figure 3 The flowchart of the adaptive control method for intelligent coupled heat exchanger units provided by this invention is shown.

[0025] Figure 4 This is a flowchart illustrating the dynamic optimization of the split ratio in the adaptive control method for intelligent coupled heat exchanger units provided by this invention.

[0026] Figure 5 The flowchart illustrates the execution of the feedforward-feedback composite control in the adaptive control method for intelligent coupled heat exchanger units provided by this invention.

[0027] Figure 6 The flowchart illustrates the antifreeze response operation in the adaptive control method for intelligent coupled heat exchanger units provided by this invention.

[0028] Figure 7 The flowchart of variable gain PID control in the adaptive control method of the intelligent coupled heat exchanger unit provided by the present invention is shown. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example 1

[0030] Existing technologies are not yet perfect in terms of structural design and control mechanisms for heat exchanger units. Traditional full-flow series connection methods have problems such as large additional resistance in the pipeline network and limited improvement in the quality of secondary side heating. Furthermore, control strategies based on fixed split ratio and reactive logic are difficult to adapt to dynamic operating conditions and exhibit control lag. As a result, when the unit undergoes deep energy-saving retrofitting, it faces consequences such as weak adaptability to existing pipeline networks, limited improvement in heating quality, and insufficient control precision.

[0031] To solve the above-mentioned technical problems, Embodiment 1 of the present invention provides an intelligent coupled heat exchanger unit, including a plate heat exchanger, a water source heat pump unit, a first bypass branch, a second bypass branch, an intelligent coupled controller, a temperature sensor, a pressure sensor, a flow meter, a variable frequency circulating pump, an electric regulating valve, a control cabinet, and an equipment base, specifically: The plate heat exchanger is used to bear the basic heat exchange load, and the design operating conditions are primary side / return water temperature of 97 / 52℃ and secondary side / return water temperature of 65 / 50℃.

[0032] The water source heat pump unit consists of an evaporator, compressor, condenser, and expansion valve forming a closed refrigerant cycle. The evaporator is connected in parallel to the primary return water line via a first bypass branch, and the condenser is connected in parallel to the secondary return water line via a second bypass branch.

[0033] The first bypass branch is opened on the primary return water main pipe and adopts a diversion pipe structure with a 30° oblique insertion angle to the main pipe. It diverts part of the primary return water into the evaporator, cools it down, and then merges back into the primary return water main pipe. Unlike the traditional scheme where the entire flow of primary return water enters the evaporator in series, the branch diversion design of this invention ensures that the pressure drop of the main flow is minimal and does not place an additional burden on the existing pipeline circulation pump.

[0034] The second bypass branch is opened on the secondary return water main, diverting a portion of the secondary return water into the condenser. After absorbing heat and heating up, the water flows into the secondary supply water main. Unlike the traditional scheme where the condenser draws water from the primary supply side, this invention connects the condenser directly to the secondary side, allowing the waste heat extracted by the heat pump to be directly used to improve the heating quality at the user end, achieving a directional energy transfer of "heat taken from the primary side and heat released to the secondary side".

[0035] The intelligent coupling controller consists of a PLC, a frequency converter, and an I / O module, and is connected to the frequency converter signals of the temperature sensor, pressure sensor, flow meter, variable frequency circulating pump, electric regulating valve, and the compressor of the water source heat pump unit, respectively, to collect temperature, pressure, and flow signals in real time and execute the coupling control algorithm.

[0036] In this invention, such as Figure 1The diagram illustrates the complete connection relationships and working fluid flow of the plate heat exchanger, water source heat pump (evaporator / compressor / condenser / expansion valve), first / second bypass branches, primary network supply and return water pipelines, and secondary network supply and return water pipelines. In operation: Primary network supply water (approximately 97°C) releases heat to the secondary network via the plate heat exchanger, resulting in a primary return water temperature of approximately 52°C. A portion of the primary return water is diverted to the evaporator side of the water source heat pump via the first bypass branch. In the evaporator, the low-temperature, low-pressure liquid refrigerant absorbs heat from this portion of the primary return water and evaporates into a gaseous state, cooling the primary return water to approximately 45°C. This cooled primary return water then flows into the primary return water main pipe, mixing with the original return water. The combined primary return water temperature is significantly lower than 52°C, achieving a large temperature difference return water. After returning from the user, a portion of the secondary return water is diverted to the condenser side of the water source heat pump via the second bypass branch. In the condenser, the high-temperature, high-pressure gaseous refrigerant releases heat to the secondary return water and condenses into a liquid state. The secondary return water is then heated and flows into the secondary supply water main, mixing with the secondary supply water at the outlet of the plate heat exchanger. The total secondary supply water temperature after mixing is higher than the outlet temperature of the plate heat exchanger when it operates alone, directly improving the heating quality and solving the problem that existing condensers draw water from the primary supply side, heat it, and then return it to the primary network, failing to directly serve secondary users and thus failing to improve the actual heating quality for users. In the evaporator, the refrigerant absorbs waste heat from the primary return water, the compressor compresses it to increase temperature and pressure, and releases heat to the secondary return water in the condenser. The expansion valve reduces the pressure and temperature, allowing the refrigerant to return to the evaporator, completing a cooling / heating cycle. The COP (heating capacity to input electrical power ratio) of the water source heat pump is 6.8-6.9.

[0037] It should be noted that, as Figure 2 As shown, both the first and second bypass branches adopt a shunt structure with an oblique insertion angle to the main pipe, preferably 30°, which can achieve: Minimizes pressure drop: Only a small portion of the flow passes through the heat pump, resulting in minimal additional resistance to the system and no impact on the hydraulic characteristics of the original heat exchanger and piping network. This solves the problem of traditional systems where the entire primary return water flows through the evaporator, significantly increasing additional resistance and potentially leading to insufficient primary-side flow when the existing circulating pump head remains unchanged, necessitating pump replacement or modification.

[0038] Maximizing temperature difference: A small flow rate through the evaporator produces a large temperature drop (about 4°C), while a small flow rate through the condenser produces a large temperature rise, ensuring that the heat pump always operates at its optimal operating point. Hydraulic decoupling: The branch lines and main lines are hydraulically independent, so the start-up, shutdown, and regulation of the heat pump do not affect the stable operation of the original plate heat exchanger. This solves the problem that the start-up and shutdown of traditional heat pumps change the overall resistance characteristics of the primary return water pipeline, directly affecting the operating conditions of the plate heat exchanger.

[0039] Low mixing resistance: When branch lines and main lines mix, the inclined pipe method has lower mixing resistance and better mixing effect.

[0040] As one possible implementation, the operating mode of the intelligent coupling controller is as follows: Conventional straight-through mode: During the early and late cold seasons or when the heat pump is not in use, the plate heat exchanger independently undertakes the entire heat exchange load, the system operates in the conventional heat exchange station mode, and the heat pump branch valves are closed.

[0041] Large temperature difference efficiency enhancement mode: During periods of severe cold or peak load, the controller automatically opens the heat pump branch valves and starts the water source heat pump based on the outdoor temperature, secondary water supply temperature setpoints, and / or a preset schedule, entering large temperature difference efficiency enhancement operation. The controller adjusts the operating frequency of the heat pump compressor in real time (via the frequency converter) and dynamically distributes the heat exchanger load and heat pump load to ensure that the system always operates within the optimal COP range.

[0042] Fail-safe mode: When the controller detects a fault in the water source heat pump (high pressure protection, overcurrent, sensor abnormality, etc.), the system automatically closes the heat pump branch valve within 5 seconds, cuts off the connection between the heat pump and the pipe network, and allows the system to seamlessly switch to the conventional direct-connect mode, where the plate heat exchanger independently undertakes the entire heating load, ensuring uninterrupted heating.

[0043] Furthermore, based on the above three operating modes, the following adaptive control method for intelligent coupled heat exchanger units is proposed to optimize and control the operating parameters of the intelligent coupled heat exchanger units, such as... Figure 3 As shown, it specifically includes the following: As a possible implementation method, addressing the problem that a fixed split ratio cannot adapt to load changes, causing the heat pump to deviate from its optimal operating condition, this embodiment is based on a fuzzy PID closed-loop control architecture. Using the instantaneous COP deviation of the heat pump and the evaporator temperature drop deviation as dual input variables, it outputs the split valve opening adjustment amount in real time through fuzzy inference, achieving dynamic self-optimization of the split ratio and ensuring that the heat pump always operates at its optimal operating point under all operating conditions. Figure 4 As shown, the specific implementation steps are as follows: Step S1: The intelligent coupling controller collects the following parameters in real time with a sampling period of 30 seconds: instantaneous COP value of the heat pump. evaporator side primary return water temperature drop condenser side secondary water supply temperature rise Heat pump load rate and primary side total flow .

[0044] Step S2: The controller compares the measured data with the preset target value and calculates the heat pump energy efficiency deviation. and evaporator temperature drop deviation : Heat pump energy efficiency deviation This reflects the direction and degree to which the heat pump's energy efficiency deviates from its optimal value; among them, The instantaneous COP value of the heat pump is obtained in real time by the intelligent coupling controller with a sampling period of 30 seconds; This represents the system's preset heat pump energy efficiency target value, which is set based on the heating performance coefficient when the water source heat pump is at its optimal operating point, and the value range is set between 6.5 and 6.9. Evaporator temperature drop deviation This reflects the quality of heat exchange on the evaporator side; among them, This indicates the temperature drop of the primary return water on the evaporator side, which is collected in real time by the intelligent coupling controller with a sampling period of 30 seconds. This indicates the system's preset target value for the primary return water temperature drop on the evaporator side. The value is set based on the target temperature drop generated by the primary return water flowing through the evaporator under the large temperature difference enhancement mode, and the value range is set between 4℃ and 7℃.

[0045] Step S3: Based on the heat pump energy efficiency deviation Deviation from evaporator temperature drop The adjustment amount of the electric regulating valve opening is used as the input variable for fuzzy inference. As the output variable, a fuzzy inference rule base is constructed, and dynamic adaptive closed-loop control is executed. The specific operation steps include: S31, Set heat pump efficiency deviation The theoretical domain of physics is Evaporator temperature drop deviation The theoretical domain of physics is Electric regulating valve opening adjustment amount The theoretical domain of physics is Before performing quantization calculations, when the actual collected deviation parameters exceed the physical domain boundary, a forced amplitude truncation is performed according to the corresponding boundary limit to ensure the operational effectiveness of the control model under extreme abnormal conditions.

[0046] Will , and The fuzzy domain is uniformly mapped to The interval is divided into seven fuzzy subsets: negative large ( ), negative ( ), negative small ( ),zero( ), Zheng Xiao ( ),middle( ), Zhengda ( ).

[0047] Extract the boundary parameters of the theoretical domain and calculate the quantization factor of the input variables. , Scale factor of output variable The calculation formula is: ; ; ; In the formula, The maximum permissible deviation limit for heat pump energy efficiency is a dimensionless parameter. The maximum permissible deviation limit for evaporator temperature drop is set, with the dimension being °C; The set limit value for single maximum opening adjustment, with dimensions of .

[0048] S32. To improve the sensitivity of control commands near the error zero point and their robustness in the boundary extreme value region, the edge fuzzy subset... and Using trapezoidal membership functions for intermediate fuzzy subsets , , , , A symmetrical triangular membership function is used. A general fuzzy variable is employed. (referring to) , or Taking the fuzzy numerical value as an example, it belongs to the intermediate fuzzy subset. The membership function expression for a symmetrical triangle is: ; In the formula, For variables Belongs to a specific fuzzy subset The membership degree is calculated and the results are distributed in Within the range; The center coordinate parameters of this fuzzy subset; This is the half-width parameter of the bottom edge of the fuzzy subset.

[0049] S33. Design an asymmetric dynamic fuzzy inference rule base. Specifically, for example: Establish containing A fuzzy inference rule base for control rules. The physical logic basis for rule design is: evaporator temperature drop deviation, which reflects the hydraulic and thermodynamic balance of the system. As the dominant variable, it determines the direction of aperture adjustment. A positive value indicates excessive temperature drop and insufficient flow rate, requiring an increase in the opening of the electric regulating valve; when... A negative value indicates excessive flow, requiring a reduction in the opening of the electric regulating valve. Simultaneously, the heat pump energy efficiency deviation... As a constraint variable, the adjustment range is dynamically corrected asymmetrically.

[0050] The correction logic is as follows: when the system's heating performance coefficient is significantly low ( When belonging to a negative subset, if the dominant variable requires an increase in flow ( If the value is positive, the output command will cause the valve to open rapidly to quickly increase the evaporation temperature; if the dominant variable requires a reduction in flow rate ( If the value is negative, then the valve closing rate will be forcibly reduced (e.g., the output will be changed from...). Revised to This prevents a sudden drop in evaporating pressure from further deteriorating the compressor's operating condition. When the coefficient of performance (COP) is excellent (…), When the subset belongs to the positive subset, a fast-closing and buffered-opening energy-saving convergence strategy is preferred. The specific parameters of the asymmetric fuzzy inference rules are shown in Table 1.

[0051] Table 1

[0052] S34. Use the centroid method to defuzzify the output fuzzy set obtained from fuzzy inference, and calculate the precise fuzzy value of the output variable. : ; In the formula, This represents the total number of fuzzy rules activated within the current control cycle. For the first The antecedent of the activated rule is comprehensively defined as its membership degree; For the first The output fuzzy subset center coordinate parameters corresponding to each rule. If the denominator approaches zero under extreme conditions, the precise fuzzy value of the previous control cycle is directly output to maintain steady state.

[0053] Precise fuzzy value Multiply by the scaling factor The final adjustment amount of the electric regulating valve opening is calculated. : ; Get Then, it is sent to the frequency converter drive module to complete the physical closed-loop adjustment of the opening of the electric regulating valve on the first bypass branch.

[0054] Step S4: The controller outputs based on fuzzy inference. The value is adjusted by driving the electric regulating valve through the frequency converter to change the opening degree. Then return to step S1 to form a closed-loop control.

[0055] Furthermore, addressing the issue of lag in response of traditional reactive control under load abrupt changes, this embodiment constructs a feedforward channel based on an outdoor temperature and load prediction model. This channel pre-sets the compressor frequency and diverter valve opening, while simultaneously superimposing incremental PID feedback correction based on real-time temperature deviation, forming a feedforward-feedback composite control. This allows the system to pre-adjust parameters before load changes occur. Figure 5 As shown, the specific implementation steps are as follows: Step F1: Access the target heat exchange station's monitoring and data acquisition system and meteorological service interface via wired or wireless communication networks to obtain multi-dimensional operating environment parameters. Construct a load forecasting model based on a long short-term memory network and establish an online fine-tuning mechanism to calculate future heat load forecasts. Specifically, this includes: Step F1-1: Extract historical operating data of the target heat exchange station over the past three complete heating seasons and set the data sampling period. for The collected feature variables include historical outdoor temperature series. Solar radiation intensity sequence Wind speed sequence Historical secondary water supply temperature sequence Compared with historical measured heating load .

[0056] The Isolation Forest algorithm was used to detect and remove outliers from the collected raw time series data, and linear interpolation was used to fill in missing values. The min-max normalization method was then used to scale the dataset to the appropriate size for all feature variables. Dimensionless intervals eliminate the interference of differences in the dimensions of different physical parameters on network weight updates.

[0057] Set the time sliding window step size parameter Construct time-aligned input tensors At the current discrete time step Based on, extract from to Historical feature set, input tensor The expression form is: .

[0058] Step F1-2: Construct a Long Short-Term Memory (LSTM) network model containing an input layer, hidden layers, and an output layer. The hidden layer is set to a two-layer structure, and the number of nodes in the first layer is set to [number missing]. The number of nodes in the second-layer network is set to Select The function is used as the hidden layer activation function to capture the nonlinear dynamic characteristics of heating load changing with multidimensional meteorological conditions. A discard rate parameter is configured between the two network layers. The node dropout layer is used to suppress model overfitting, and the output layer uses a single-node fully connected layer and The activation function outputs the normalized prediction result.

[0059] Define a comprehensive loss function that incorporates the physical constraints of building thermal inertia. Used for offline model training. Since the input data has been normalized to... Within the interval, all components of the loss function are dimensionless, ensuring dimensional consistency during the calculation process. The formula for calculating the comprehensive loss function is: ; in, The mean squared error term is used to measure prediction bias, and its calculation formula is: ; In the formula, This refers to the batch size of the training samples. For the first Normalized predicted load values ​​for each sample; For the first Normalized measured load values ​​for each sample.

[0060] The physical constraint penalty term is constructed based on the maximum rate of change of thermal inertia of the pipeline network and building, abandoning the single temperature monotonicity constraint. The calculation formula is as follows: ; In the formula, The preset normalized threshold for the maximum allowable load change rate between adjacent time steps is calibrated using the physical thermal inertia time constant of the heating network to limit abnormal step jumps in the output curve that violate objective physical laws. Penalty weighting coefficient. Set as .

[0061] use The optimizer sets the initial learning rate to be... The model is iteratively updated until the comprehensive loss function is continuous. The wheel's descent was less than The early stopping mechanism is triggered to complete offline training and solidify the basic weight parameters of the network.

[0062] Steps F1-3: Obtain the future weather forecast data for the area where the target heat exchange station is located through the meteorological bureau's numerical weather prediction application interface. to Hourly gridded weather forecast data, including outdoor temperature forecast, solar radiation forecast, and wind speed forecast. The spatial resolution of the numerical weather prediction is [missing information]. The time resolution is The cubic spline interpolation algorithm is used to... The resolution forecast data is downscaled to a step size of . Continuous forecast sequences.

[0063] The system's current-time feature tensor is extracted and combined with the transformed future weather forecast sequence. After normalization, it is input into the load prediction model for forward propagation calculation. The numerical results output by the network are then inversely normalized to generate a model with... Dimensional future heat load forecast ,in To predict the leading time span.

[0064] Steps F1-4: As the heating season progresses and equipment ages physically, activate the online adaptive update mechanism for the predictive model. The controller... The day serves as a rolling evaluation cycle, calculating past... Actual heating load during the day Compared with historical forecast load The root mean square error between them.

[0065] When the root mean square error is continuous Hours exceeding the set threshold When this occurs, an online fine-tuning command is triggered. The controller retrieves the most recent... An incremental training set is generated from the actual operating data of the network. The parameters of the first hidden layer of the Long Short-Term Memory network are frozen, while only the parameters of the second hidden layer and the output layer are unfrozen. A decaying learning rate is then applied. Perform incremental iterative training. Once the prediction error on the incremental training dataset converges to within the threshold, update the weights of the underlying operating model of the controller to eliminate the data distribution offset caused by pipeline resistance drift and heat exchanger fouling.

[0066] Step F2: Based on future heat load forecasts The controller pre-sets the operating frequency of the heat pump compressor. and the opening degree of the diversion valve This ensures that the heat pump adjusts to the expected operating point before load changes occur. The specific implementation steps are as follows: F2-1, The controller extracts the heat pump load sharing ratio generated by the dynamic optimization of the system. Calculate the preset heating capacity of the heat pump at future times. According to the calculated Match the fitting equation of the compressor's variable frequency characteristic and calculate the operating frequency of the heat pump compressor. : In the equation , , These are the characteristic coefficients fitted to the historical steady-state operating data of the compressor.

[0067] F2-2. Calculate the target split flow rate required to extract low-grade heat from the evaporator side based on the thermodynamic law of conservation of energy. Furthermore, to eliminate the calculation error in heat exchange caused by the actual operating energy efficiency of the heat pump deviating from the ideal setting, the effective heating performance coefficient of the heat pump, after smoothing, is used instead of the static energy efficiency target value. The calculation formula is as follows: ; In the formula, Traffic is diverted to target future moments, with the dimension being... ; The heat pump is pre-set to provide heating capacity for future use, with dimensions of [missing information]. ; The effective heating performance coefficient of the heat pump at the current moment is dimensionless; The specific heat capacity of circulating water at constant pressure, with dimensions of ; The density of the circulating water has dimensions of . ; The preset target value for the primary return water temperature drop on the evaporator side is expressed in °C. This is the conversion factor from seconds to hours for time units.

[0068] Among them, in calculating the effective heating performance coefficient of the heat pump at the current moment At that time, a first-order low-pass filtering algorithm was used to analyze the instantaneous COP value of the heat pump acquired in real time. The transient measurement noise is filtered out, and the calculation formula is as follows: ; In the formula, These are the filter weight coefficients, with a value range set to... .

[0069] To ensure the effectiveness of the physical model under extreme conditions, the controller... Execute absolute lower bound constraint interception. This occurs when the heat pump is in the start-stop transition phase or a throttling failure occurs. At that time, a mandatory order This prevents non-physical negative or zero values ​​from being calculated for the flow target value, ensuring the stability of the feedforward regulation path.

[0070] F2-3. It is clarified that the electric regulating valve on the first bypass branch uses a valve core with an equal percentage flow characteristic. This characteristic can compensate for the nonlinear heat transfer characteristics of the evaporator under different operating conditions, keeping the overall amplification factor of the control channel constant. The controller retrieves the inverse mapping function of the flow characteristic of this electric regulating valve to calculate the preset opening degree of the diversion valve at future times. Furthermore, to ensure the continuity and physical effectiveness of the control model across the entire flow range, and to eliminate the mathematical singularity of the equal percentage function near zero, a piecewise calculation function incorporating boundary constraint logic is constructed: when When the target flow rate is below the valve's effective control limit, the output is forcibly cut off. ; when When the condition is within the effective adjustment range, the calculation is performed using the inverse mapping function: ; when When the target flow rate has reached the physical full load boundary, the output is forcibly truncated. ; In the formula, The pre-set opening degree of the diverter valve for a future time is a dimensionless parameter, and its value range is as follows: ; Traffic is diverted to target future moments, with the dimension being... ; The maximum rated flow rate when the electric regulating valve is fully open, with dimensions of ; This is the inherent adjustable ratio parameter calibrated at the factory for electric control valves. It is dimensionless and its physical meaning is the ratio of the maximum flow rate to the minimum flow rate that the valve can control under a specified pressure drop.

[0071] During the system parameter configuration phase, read the calibration parameters from the nameplate of the electric regulating valve and set the inherent adjustable ratio. The exact value. For standard equal percentage valves. The range of values ​​is in to Between these values, the preferred typical value in this embodiment is... Load the underlying registers. Through piecewise function design, ensure that the split instruction can output safe absolute boundary values ​​under both minimal and overload conditions.

[0072] Step F3: The controller uses a fixed sampling period. Real-time acquisition of the current discrete time step The actual temperature of the secondary water supply Calculate its actual value and the actual temperature setpoint. Temperature deviation between the current time : The feedback correction amount is calculated using an incremental PID algorithm. The specific implementation steps are as follows: F3-1, The controller retrieves the temperature deviation from the previous time step stored in the register. Temperature deviation from the previous two time steps And retrieve the proportional gain coefficient that matches the current operating condition range. Integral gain coefficient and differential gain coefficient .

[0073] F3-2. Based on the extracted discrete error sequence, the frequency feedback correction is calculated using an incremental PID algorithm. The calculation formula is: .

[0074] F3-3. Introduce limiting dead-zone logic to eliminate measurement noise and prevent frequent compressor operation; set a temperature control dead-zone threshold. Adjustment step size of maximum permissible frequency in a single operation , judge when At that time, the output is forcibly truncated. ; Determine when the calculated result is At that time, the hard limit command is executed and output. The generated safety feedback correction quantity, which conforms to physical constraints, is then fed into the composite output stage.

[0075] Step F4: Calculate the final operating frequency of the heat pump compressor. : This enables composite control combining feedforward prediction and feedback correction.

[0076] This method can shorten the system response time to less than 5 minutes during sudden load changes, and reduce the fluctuation range of secondary water supply temperature from... ℃ decreased to ℃, thereby improving the heating comfort of users.

[0077] In this embodiment, the heat pump load sharing ratio The value of is dynamically optimized based on the principle of minimizing the overall energy consumption of the system. A cost function is constructed with the objective of minimizing the total input electrical power of the system under predicted heat load: In the formula, The total input electrical power of the system, with dimensions of ; The predicted heat load for future times, with dimensions of ; The expected heating performance coefficient of the heat pump under the corresponding load sharing ratio is dimensionless. The expected power consumption of the variable frequency circulating pump in the pipeline network under the corresponding sharing ratio is expressed in units of... Among them, the expected coefficient of performance (COP) The calculation is based on the heat pump partial load factor. The polynomial fitting characteristic curve is determined as follows: , In the formula, The rated thermal performance coefficient specified on the nameplate of a water source heat pump is dimensionless. The unit of heat is 1000 kilocalories, with the dimension 1000. ; , , These are the dimensionless characteristic coefficients obtained through offline fitting based on historical steady-state operating data. The cost function is minimized. To find the optimal target, decision variables are set in combination with the physical limits of the equipment. Feasible region constraints: In the formula, The minimum load sharing ratio threshold is used to ensure stable operation of the heat pump compressor. This threshold is set when the system is under extremely low load and the calculated upper limit of the feasible region is less than [value missing]. Force output Solving the problem within the defined feasible region using a nonlinear optimization algorithm makes... The minimum global optimal solution is used as the target heat pump load sharing ratio.

[0078] Compressor frequency conversion characteristic fitting coefficient , , The values ​​are obtained offline through least squares analysis based on the historical steady-state operation dataset of the heat pump, and are collected from the data stored in the controller. Data on heat pump heating capacity and corresponding compressor operating frequency under historical steady-state conditions. Construct the output column vector With the input feature matrix According to the analytical solution formula The calculations are then stored in the controller's underlying registers.

[0079] Specific heat capacity of circulating water at constant pressure The value is determined by polynomial mapping calculation based on the characteristic temperature of the thermodynamic state of the working fluid, using the real-time collected primary return water inlet temperature on the evaporator side. Substitute the standard water property fitting correlation into the input variable. The calculation yields a value in kJ / (kg·℃).

[0080] Actual temperature setpoint The determination is based on the climate compensation logic established by the building heat balance principle, which is calculated using feedforward calculations and by obtaining real-time outdoor temperatures from meteorological services. Retrieve the outdoor minimum temperature preset for heating design at the heat exchange station. Design outdoor maximum temperature The corresponding upper limit temperature of secondary water supply and the lower limit temperature of secondary water supply According to the formula Calculate the actual temperature setpoint.

[0081] Temperature control dead zone threshold The value is determined based on the hardware measurement uncertainty calibration of the secondary water supply temperature sensor, by extracting the nominal absolute accuracy error value of the temperature sensor from its factory specifications. Noise attenuation coefficient of anti-aliasing filter in signal acquisition module According to the formula Calculate the fixed dead zone span to eliminate erroneous adjustment fluctuations caused by instrument white noise.

[0082] Maximum permissible frequency adjustment step size per cycle The value is determined based on the hardware electrical safety speed limit parameters of the compressor inverter, and is obtained by extracting the maximum frequency change rate hard parameter limited in the inverter's manufacturer's manual. With the fixed control sampling period set by the controller's operating program According to the formula Define the absolute safety boundary for single-step physical adjustment and determine the maximum permissible frequency adjustment step size for a single operation.

[0083] Furthermore, addressing the issue of evaporator freezing and cracking due to low-temperature bypass return water under extremely cold conditions, this embodiment utilizes a three-level antifreeze threshold monitoring and evaporator full bypass switching mechanism. It executes flow ratio reduction and evaporator full bypass operation at the warning and danger levels respectively, and adaptively resumes forward operation after the risk is eliminated, achieving unattended safe antifreeze protection. Figure 6 As shown, the specific implementation steps are as follows: Step P1: Distribute evenly along the circumference of the evaporator outlet pipe. Each sensor is independent. An asymmetric data fusion strategy, designed with safety in mind, is used to output the evaporator outlet temperature. The specific operation is as follows: Remove deviations The average real-time readings of the sensors exceed the set deviation threshold (e.g., The abnormal fault readings are extracted, and the minimum value is extracted from the remaining valid readings as the output result. This eliminates the hidden danger of readings being too high due to uneven pipe wall temperature distribution and ensures that the antifreeze response is triggered as soon as local extremely low temperatures occur.

[0084] Setting a dynamic antifreeze threshold: Obtaining the physical freezing point temperature of the circulating working fluid in the heat exchange system. Combined with the evaporator design, the minimum heat transfer temperature difference Calculate the antifreeze thresholds for each level. Warning level threshold. Hazard level threshold Recovery level threshold ,in and This is the preset absolute temperature safety margin constant, and .

[0085] Step P2, when When this occurs, a warning response is triggered. The controller determines that there is excessive heat extraction on the evaporator side or insufficient return water flow, and sends a control command to the frequency converter and the electric regulating valve: increase the current opening of the electric regulating valve on the first bypass branch by the set absolute opening value. The water-side heat exchange temperature drop is reduced by increasing the return water flow rate into the evaporator; simultaneously, the operating frequency of the heat pump compressor is lowered by the set absolute frequency value. This reduces the refrigerant circulation rate and heat absorption, thus suppressing further decreases in evaporation temperature.

[0086] Step P3: Regardless of the current operating state of the system, as long as Immediately trigger the hazardous state response operation. The controller prioritizes sending a shutdown command to shut down the heat pump compressor, cutting off the system's internal heat extraction mechanism; simultaneously, it closes the electric regulating valve on the first bypass branch and opens the bypass short-circuit valve, allowing the circulating water remaining in the first bypass branch and evaporator tube bundle to form a local self-circulation flow in the closed short-circuit circulation loop formed by the evaporator and the bypass short-circuit valve, relying on the heat pump unit's own circulation pump. This utilizes the kinetic energy of water movement to prevent static water freezing, and outputs a system low-temperature alarm signal. During the temperature recovery process, as long as... The controller locks the dangerous state response mode, prohibits premature degradation of the control state, and eliminates command oscillations caused by measurement fluctuations.

[0087] Step P4, when When the duration reaches the set time threshold, the freezing risk is deemed relieved, the status lock is released, and a step-by-step recovery procedure is initiated. First, the bypass short-circuit valve is closed. Then, according to the set time intervals and absolute opening step sizes, the opening of the electric regulating valve on the first bypass branch is gradually increased. Once the water flow in the first bypass branch recovers to the safe operating lower limit and the water temperature stabilizes, the heat pump compressor is restarted, and the operating frequency is gradually restored according to the set absolute frequency step sizes. During the step-by-step recovery process, an antifreeze threshold interruption callback closed-loop monitoring mechanism is executed: if... It fell back to the warning level threshold. Immediately interrupt the recovery procedure and roll back to step P2; if Falling to the danger level threshold Immediately interrupt the recovery procedure and roll back to step P3 until all system status parameters fully converge and the system returns to its normal operating state before the alert response operation was triggered.

[0088] Step P5, when If the system is deemed to have no risk of evaporator icing, it maintains normal operation. The controller keeps the bypass short-circuit valve closed and releases the antifreeze mechanism from locking the highest priority commands of the underlying actuators. The intelligent coupling controller fully delegates control to the flow ratio dynamic self-optimization channel, the feedforward-feedback composite control channel, and the variable gain PID control channel. Based on the instantaneous COP value of the heat pump, the temperature drop of the primary return water on the evaporator side, the actual temperature of the secondary supply water, and outdoor weather forecast data, it calculates and dynamically issues the opening degree of the electric regulating valve of the first bypass branch and the operating frequency of the heat pump compressor in real time, ensuring that the water source heat pump unit performs routine optimization control to maximize system energy efficiency and heating quality within the safety boundary of no icing risk.

[0089] This method can effectively prevent evaporator freezing under extreme conditions of -20℃, enabling unattended safe operation of the heat exchange station and reducing unplanned downtime compared to conventional shutdown antifreeze methods.

[0090] Furthermore, addressing the issue of accurately locking the secondary water supply temperature under multiple disturbances, this embodiment employs a segmented variable-gain PID control architecture based on the secondary return water temperature. Independent PID parameters are configured for low-load, normal-load, and high-load zones, and parameter smoothing is performed during zone transitions to achieve high-precision locking of the secondary water supply temperature across all operating conditions. Figure 7 As shown, the specific implementation steps are as follows: Step C1: Set the low load zone threshold and high load zone threshold The controller is based on the secondary return water temperature. The operating conditions are divided into three intervals: Interval A setting This is a low-load area; Interval B setting This is the normal load zone; Interval C setting This is a high-load area.

[0091] In this embodiment, the low load zone threshold The narrow-point temperature difference is defined based on the design of the plate heat exchanger. Specifically: the design temperature of the primary side return water within the system is extracted. With secondary side return water design temperature It can be calculated using the following formula: The logic of this formula lies in defining the difference between the primary side return water design temperature and the secondary side return water design temperature as the design narrow point temperature difference of the heat exchanger unit. When the secondary return water temperature is higher than the design value, and its deviation reaches this narrow point temperature difference, the heat transfer driving force of the plate heat exchanger is significantly reduced, indicating that the system is in a low-load zone with minimal heat consumption. Preferably, the following is set... The temperature is 48℃.

[0092] In this embodiment, the high load zone threshold The coupling boundary between the supply and return water temperature difference in the pipeline network design and the maximum heat extraction capacity of the heat pump is defined. Specifically: the design temperature of the secondary side supply water built into the extraction system. and the system's preset target maximum value for primary return water temperature drop on the evaporator side. It can be calculated using the following formula: The formula's logic lies in calculating the design temperature difference of the secondary side piping network (i.e., the secondary side supply water design temperature minus the secondary side return water design temperature), and subtracting the maximum allowable temperature drop of the heat pump evaporator to obtain the system's residual limit temperature drop range. When the secondary return water temperature falls below this residual range compared to the design value, it indicates that the combined heating capacity of the plate heat exchanger and heat pump unit has reached the full-load physical boundary, and the system enters the high-load zone. Preferably, a setting is provided. The temperature is 42℃.

[0093] Step C2: Invoke the proportional gain coefficient, integral gain coefficient, and derivative gain coefficient that match the operating condition range. Specifically: The initial gain coefficients for each operating condition range are established as reference parameters through the relay feedback self-tuning test performed on the heat exchanger unit during the on-site commissioning phase, and the critical oscillation gain of the system is extracted. With critical oscillation period The Ziegler-Nichols rule was used to quantitatively calculate the baseline proportional gain coefficient. Reference integral gain coefficient Differential gain coefficient with reference The calculation formula is: , , ; In the formula, The dimensions are ; The dimensions are ; The dimensions are ; The dimensions are ; The dimensions are To accommodate the thermodynamic nonlinear characteristics under different operating conditions, dimensionless dynamic weighting coefficients are introduced across the operating condition range to correct the reference parameters. Independent proportional gain coefficients, integral gain coefficients, and differential gain coefficients are configured for the low-load, normal-load, and high-load regions, respectively. The formula for calculating the gain coefficient is as follows: , , ; In the formula, These represent the low-load zone, normal-load zone, and high-load zone, respectively. , , These are the dimensionless weighted coefficients for the proportional, integral, and differential equations, respectively, under the corresponding operating condition intervals. Specifically, for example: intervals... Low load zone setting , , Temperature overshoot is suppressed by reducing the proportional and integral actions and enhancing the derivative action; range Normal load zone setting , , The response speed and stability of the control system are balanced using reference parameters; interval High load zone setting , , By increasing the proportional and integral effects, the response gain is improved, which can compensate for the regulation lag caused by the increased thermal inertia of the heat exchange unit during the cold season.

[0094] Step C3, when When crossing the interval boundary point, the PID parameters gradually change to the target value linearly within 10 seconds to prevent control oscillations caused by sudden parameter changes.

[0095] The gradual change process is as follows: ,in This is the number of seconds since the moment of travel.

[0096] Step C4: Use the secondary water supply temperature setpoint To control the target, the actual secondary water supply temperature is used. As the controlled variable, the frequency regulation of the heat pump compressor is calculated using an incremental PID algorithm. By adjusting the compressor speed through a frequency converter, the secondary water supply temperature can be locked with high precision.

[0097] This method can control the fluctuation range of secondary water supply temperature within a certain range across the entire load range. Within ℃, thereby improving control accuracy.

[0098] The four control mechanisms described above operate collaboratively within the intelligent coupled controller, sharing the same set of temperature, pressure, and flow sensor data, and the same execution channel of the heat pump compressor inverter. Each control mechanism coordinates its actions through a priority arbitration mechanism, collectively forming an adaptive control system for the intelligent coupled heat exchanger unit. Among these, the antifreeze protection mechanism has the highest priority and can unconditionally override the control commands for the diversion valve and compressor. Under normal operating conditions, the compressor frequency command output by the feedforward-feedback composite control mechanism is superimposed with the frequency adjustment output by the variable gain PID control mechanism before execution. The diversion ratio optimization mechanism independently adjusts the diversion valve opening, decoupled from the temperature control channel.

[0099] As an optional embodiment, to verify the physical effectiveness and system stability of the intelligent coupled heat exchanger unit and its adaptive control method under different operating boundary conditions, a comparative verification scheme including a control group and an experimental group is established. The control group adopts a constant split ratio of existing technology (setting the actual split ratio). The experimental group adopted the adaptive control method proposed in this invention, which introduces a dynamic self-optimization of the split ratio and a feedforward-feedback composite control mechanism based on fuzzy inference.

[0100] Three representative ambient temperature and load ranges were selected, and continuous [testing / implementation] was performed. Steady-state and dynamic disturbance tests were conducted. The test conditions were based on the primary water supply temperature. With secondary water supply set temperature The differences are categorized into low-load, standard-load, and high-load operating conditions. The test dataset includes the actual split ratio of the heat pump. Variation range, average energy efficiency Energy efficiency standard deviation and the standard deviation of secondary water supply temperature The specific steady-state operating performance comparison data is shown in Table 2.

[0101] Table 2

[0102] Introducing an amplitude during the unit's steady-state operation A load step disturbance was detected, and a dynamic response test of the system was performed. The test module recorded the system's dynamic response in real time from the moment the disturbance occurred until the actual temperature of the secondary water supply stabilized at the set value. Required response time within the error band Maximum temperature overshoot during the transition process and the opening degree and frequency of the electric regulating valve. The specific dynamic response test data is shown in Table 3.

[0103] Table 3

[0104] Steady-state test data shows that the heating performance coefficient of the water source heat pump unit is constrained by the physical boundary of the reverse Carnot cycle, and varies with the set temperature of the secondary water supply on the condenser side. The increase in efficiency showed a deterministic decreasing trend. In the control group, due to the fixed split ratio setting, a physical mismatch occurred between the low-grade heat extraction on the evaporator side and the heat output demand on the condenser side when the system deviated from its design operating point, leading to an increased standard deviation in energy efficiency. As the flow rate increases, average energy efficiency is suppressed. The experimental group uses a fuzzy inference rule base to calculate the opening adjustment in real time, driving the diversion valve to dynamically optimize the actual diversion ratio. Maintaining the target heat transfer temperature drop during operating condition boundary drift ensures that the average energy efficiency across the entire operating range is maintained. Compared with the control group The efficiency is improved, and the high operational stability is maintained. ).

[0105] Dynamic test data shows that conventional reactive control logic suffers from regulation hysteresis due to thermal inertia, leading to severe temperature overshoot and increased actuator wear. The experimental group relied on a load prediction model to pre-calculate feedforward actions, combined with adaptive variable gain feedback correction under operating conditions, to reduce the response time under load step disturbances. Compress to Maximum temperature overshoot Inhibition in Regulating valve opening frequency reduce The above data verifies the physical rationality and engineering effectiveness of this control scheme under complex nonlinear conditions in the heating network.

[0106] As an optional implementation, the angle of the branch line can be flexibly adjusted according to the site space and pipeline layout, and the branch function can be realized in the range of 10°-60°, not limited to 30°.

[0107] As an alternative implementation: bypass branch diversion control can be achieved via a manual or electric regulating valve, and heat pump start-up and shutdown can also be manually switched.

[0108] As an alternative implementation: the water source heat pump can use a scroll, screw, or centrifugal compressor, and different types can be selected according to the unit capacity and operating conditions.

[0109] As an optional embodiment, the control method of the present invention is not only applicable to urban centralized heating systems, but can also be extended to scenarios such as industrial waste heat recovery, regional energy stations, and geothermal cascade utilization.

[0110] As an alternative embodiment, the first bypass branch and the second bypass branch can also be connected to different heat pump units, that is, the evaporator side and the condenser side each use independent refrigeration cycles, instead of sharing a single heat pump.

[0111] The preferred embodiments of the invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. An adaptive control method for an intelligent coupled heat exchanger unit, characterized in that, include: Obtain the instantaneous COP value of the heat pump of the heat exchange unit, the temperature drop of the primary return water on the evaporator side, the outdoor temperature prediction curve, the actual temperature of the secondary water supply, the evaporator outlet temperature and the secondary return water temperature; Based on the instantaneous COP value of the heat pump and the temperature drop of the primary return water on the evaporator side, the heat pump energy efficiency deviation and the evaporator temperature drop deviation are calculated. The opening adjustment amount of the electric regulating valve is determined by fuzzy reasoning. The opening is adjusted according to the opening adjustment amount, and the system returns to continue to obtain the instantaneous COP value of the heat pump and the temperature drop of the primary return water on the evaporator side to form a closed-loop control. The system acquires outdoor temperature forecast, solar radiation forecast, and wind speed forecast, determines future heat load forecast, and performs feedback correction based on the temperature deviation between the actual secondary water supply temperature and the actual temperature setpoint to determine the final operating frequency of the heat pump compressor. Based on the evaporator outlet temperature, a corresponding antifreeze response operation is triggered to control the electric regulating valve and bypass short-circuit valve on the first bypass branch. Based on the secondary return water temperature, the operating range is divided, and the operating frequency of the heat pump compressor is adjusted by calling the matching proportional gain coefficient, integral gain coefficient, and derivative gain coefficient.

2. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 1, characterized in that, Based on the instantaneous COP value of the heat pump and the temperature drop of the primary return water on the evaporator side, the heat pump efficiency deviation and the evaporator temperature drop deviation are calculated. The adjustment amount of the electric regulating valve opening is determined through fuzzy reasoning, including: The difference between the instantaneous COP value of the heat pump and the preset heat pump energy efficiency target value is used to obtain the heat pump energy efficiency deviation. The temperature drop deviation of the evaporator is obtained by subtracting the temperature drop of the primary return water on the evaporator side from the preset target value of the temperature drop of the primary return water on the evaporator side. Calculate the quantization factors of the heat pump energy efficiency deviation and the evaporator temperature drop deviation, as well as the proportional factor of the electric regulating valve opening adjustment amount. Map the physical domains of the heat pump energy efficiency deviation, the evaporator temperature drop deviation and the electric regulating valve opening adjustment amount to a fuzzy domain interval and divide them into multiple fuzzy subsets. Using the heat pump energy efficiency deviation and evaporator temperature drop deviation as input variables for fuzzy inference, the precise fuzzy value of the electric regulating valve opening adjustment amount is output according to the fuzzy inference rule base. The precise fuzzy value is multiplied by the proportional factor to calculate the electric regulating valve opening adjustment amount.

3. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 2, characterized in that, The precise fuzzy value of the electric regulating valve opening adjustment amount is output based on the fuzzy inference rule base, including: A fuzzy inference rule base containing multiple control rules is established. The evaporator temperature drop deviation, which reflects the hydraulic and thermodynamic balance of the system, is used as the dominant variable to determine the direction of the opening adjustment. The heat pump energy efficiency deviation is used as the constraint variable to perform asymmetric dynamic correction of the adjustment range. When the heat pump energy efficiency deviation belongs to the negative subset, if the evaporator temperature drop deviation is positive, a command is output to accelerate the opening of the electric regulating valve; if the evaporator temperature drop deviation is negative, the closing rate of the electric regulating valve is forcibly reduced. When the heat pump energy efficiency deviation belongs to a positive subset, a convergence strategy of rapid shutdown and buffered startup is adopted. The centroid method is used to defuzzify the output fuzzy set obtained from fuzzy inference, and the precise fuzzy value of the opening adjustment amount of the electric regulating valve is calculated.

4. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 1, characterized in that, Determining future heat load forecasts based on the outdoor temperature prediction curve includes: Historical outdoor temperature sequences, historical secondary water supply temperature sequences, and historical measured heating loads were collected as feature variables and packaged into a time series input tensor using the sliding window method. The time series input tensor is fed into a long short-term memory network for processing. The network is then trained offline using a comprehensive loss function that incorporates physical constraints related to building thermal inertia, and the predicted future heat load is output. The comprehensive loss function includes a mean square error term and a physical constraint penalty term constructed based on the maximum allowable rate of change of load between adjacent time steps.

5. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 4, characterized in that, Determining the operating frequency of the heat pump compressor and the opening degree of the diverter valve based on the predicted future heat load includes: A cost function is constructed with the goal of minimizing the total input power of the system under the predicted heat load. Under the constraints of the set feasible region, the target heat pump load sharing ratio is extracted by optimization. The preset heating capacity of the heat pump at future time is determined based on the product of the target heat pump load sharing ratio and the predicted value of the future heat load. Substitute the preset heating capacity of the heat pump at future time into the fitting equation of the compressor's frequency conversion characteristics to determine the operating frequency of the heat pump compressor. The instantaneous COP value of the heat pump collected in real time is processed by a first-order low-pass filter algorithm to calculate the effective heating performance coefficient of the heat pump at the current moment. The target diversion flow rate is determined by combining the preset heating capacity of the heat pump at the future moment, the effective heating performance coefficient of the heat pump at the current moment, the specific heat capacity of the circulating water at constant pressure, the density of the circulating water, and the preset target value of the primary return water temperature drop on the evaporator side. The opening degree of the diversion valve on the first bypass branch is determined by using the inverse mapping function between the target diversion flow rate and the equal percentage flow characteristic of the electric regulating valve, combined with the piecewise calculation function of the boundary constraint logic.

6. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 5, characterized in that, The determination of the final operating frequency of the heat pump compressor includes: Extract the temperature deviation of the previous time step and the temperature deviation of the previous two time steps, and combine the temperature deviation, proportional gain coefficient, integral gain coefficient and derivative gain coefficient to calculate the frequency feedback correction using an incremental PID algorithm. The frequency feedback correction is limited based on the set temperature control dead zone threshold and the maximum allowable frequency adjustment step size in a single operation. The final operating frequency of the heat pump compressor is determined by adding the operating frequency of the heat pump compressor to the frequency feedback correction amount after amplitude limiting.

7. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 1, characterized in that, Based on the evaporator outlet temperature, a corresponding antifreeze response operation is triggered, including: Collect the evaporator outlet temperature and set warning level threshold, danger level threshold and recovery level threshold; When the evaporator outlet temperature exceeds the warning level threshold, normal operation is maintained. When the evaporator outlet temperature is greater than the danger level threshold and less than or equal to the warning level threshold, a warning state response operation is triggered, which increases the opening of the electric regulating valve on the first bypass branch and decreases the operating frequency of the heat pump compressor. When the evaporator outlet temperature is less than or equal to the danger level threshold, a danger state response operation is forcibly triggered, the heat pump compressor and the electric regulating valve are shut down, the bypass short-circuit valve is opened to establish internal water self-circulation, and the danger state response operation is locked while the evaporator outlet temperature is less than the recovery level threshold. When the evaporator outlet temperature is greater than or equal to the recovery stage threshold and continues for a set time threshold, the state lock is released and the step-by-step recovery program is started. The bypass short-circuit valve is closed and the opening of the electric regulating valve is increased step by step. After the water temperature stabilizes, the heat pump compressor is restarted and the operating frequency is restored step by step. Closed-loop monitoring is performed during the execution of the step-by-step recovery procedure. If the evaporator outlet temperature drops to the warning level threshold or the danger level threshold, the step-by-step recovery procedure is immediately interrupted and the corresponding antifreeze response operation is rolled back.

8. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 6, characterized in that, Based on the secondary return water temperature, the operating range is divided, and the operating frequency of the heat pump compressor is adjusted by calling the matching proportional gain coefficient, integral gain coefficient, and derivative gain coefficient, including: Based on the set low load zone threshold and high load zone threshold, the operating conditions are divided into low load zone, normal load zone and high load zone according to the secondary return water temperature, and independent proportional gain coefficient, integral gain coefficient and differential gain coefficient are configured for low load zone, normal load zone and high load zone respectively. Determine the operating range to which the secondary return water temperature belongs, and extract the corresponding proportional gain coefficient, integral gain coefficient, and differential gain coefficient. The setpoint of the secondary water supply temperature is used as the control target, and the actual temperature of the secondary water supply is used as the controlled variable. An incremental PID algorithm is used to calculate the frequency adjustment of the heat pump compressor.

9. The adaptive control method for the intelligent coupled heat exchanger unit according to claim 8, characterized in that, The process of extracting the corresponding proportional gain coefficient, integral gain coefficient, and differential gain coefficient also includes: when the secondary return water temperature crosses the boundary point of the operating condition range, the proportional gain coefficient, integral gain coefficient, and differential gain coefficient perform linear gradual change within a set gradual change time based on the difference between the current coefficient value and the target coefficient value.

10. An intelligent coupled heat exchanger unit, employing the adaptive control method according to any one of claims 1 to 9, characterized in that, include: Plate heat exchanger, water source heat pump unit, first bypass branch, second bypass branch, intelligent coupling controller, temperature sensor, pressure sensor, flow meter, variable frequency circulating pump and electric regulating valve; The water source heat pump unit includes an evaporator, a compressor, a condenser, and an expansion valve; The evaporator is connected in parallel to the primary return water pipeline via a first bypass branch, and the condenser is connected in parallel to the secondary return water pipeline via a second bypass branch. Both the first and second bypass branches adopt a branch pipe structure that is inclined to the main pipe. The intelligent coupling controller is connected to the temperature sensor, pressure sensor, flow meter, variable frequency circulating pump, electric regulating valve, and compressor frequency converter of the water source heat pump unit.

Images