An air source heat pump system with adaptive load regulation and control method

Abstract

The application relates to the technical field of heat pump control, and discloses an air source heat pump system with self-adaptive load regulation and a control method.The system comprises a physical execution layer for executing a thermodynamic cycle; a state sensing module for collecting physical parameters in real time and calculating an actual thermodynamic state vector; an optimization path database for storing a pre-generated optimal thermodynamic state path and a corresponding time sequence feedforward control sequence; a transient path planning module for matching and calling the optimal path and the time sequence feedforward control sequence based on the actual thermodynamic state vector in response to a target load and a response time limit; and a path tracking execution module for receiving the path and the feedforward sequence, generating a correction control signal based on a state deviation vector, superimposing the correction control signal and the feedforward sequence, generating a final time sequence control instruction and issuing the final time sequence control instruction.The application realizes response time limit self-adaptation under low computing power by means of the optimization path database and the transient path planning, and takes into account transient energy efficiency and regulation speed.

Description

Technical Field

[0001] This invention relates to the field of heat pump control technology, specifically to an adaptive load-regulating air source heat pump system and control method. Background Technology

[0002] Air source heat pump systems, as highly efficient heating and cooling devices, have been widely used in building energy conservation. In actual operation, heat pump systems frequently need to respond to changes in the outdoor environment and fluctuations in indoor user demand, requiring the control system to perform rapid and efficient transient load adjustments. Therefore, maximizing the system's energy utilization efficiency during dynamic adjustment while ensuring user comfort is a core technical issue that continues to be a focus in this field.

[0003] To achieve the aforementioned adjustments, existing technologies primarily employ setpoint tracking-based control strategies. A common approach is to use a proportional-integral-derivative (PID) controller, which adjusts actuators such as compressor frequency and expansion valve opening by detecting the deviation between actual parameters (e.g., outlet water temperature or indoor temperature) and the setpoint. Another approach is to employ model-based advanced control, such as model predictive control (MPC). This approach attempts to establish a thermodynamic model of the system and calculate control commands in real-time by solving an optimization problem online, aiming to minimize energy consumption while satisfying constraints.

[0004] While existing technologies (such as PID control or MPC control) have achieved automatic adjustment to some extent, some shortcomings still exist:

[0005] First, traditional PID controllers rely on fixed parameter tuning, making it difficult to balance speed and economy in adjustment. Once the parameters are determined, the control strategy cannot be dynamically adjusted according to the user's real-time needs. Second, PID control logic passively responds to deviations, lacking global optimization for transient processes. When the system has high inertia or encounters external disturbances, it is prone to overshoot, oscillation, or response hysteresis, leading to additional energy losses. For advanced solutions such as model predictive control, real-time, highly complex nonlinear optimization calculations are required on resource-constrained embedded controllers. This places extremely high demands on the controller's computing power, resulting in high engineering costs and difficult deployment. Furthermore, neither PID nor conventional MPC control models can effectively handle highly nonlinear and discontinuous special operating conditions such as defrosting start-stop. When the system switches between different operating modes, conventional control logic often fails, leading to reduced system reliability and additional energy consumption. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an adaptive load-regulating air source heat pump system and control method, which solves the problems that existing heat pump control technologies struggle to balance rapid response and operational economy when performing transient load regulation, and that advanced algorithms require excessive online computing power.

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

[0008] The first aspect of this invention provides an adaptive load-regulating air-source heat pump system, the system comprising:

[0009] The physical execution layer is used to execute thermodynamic cycles;

[0010] A state-aware module, which is connected to the physical execution layer, is used to collect physical parameters in real time and calculate and generate the actual thermodynamic state vector.

[0011] An optimized path database is used to store pre-generated optimal thermodynamic state paths and time-series feedforward control sequences corresponding to the optimal thermodynamic state paths.

[0012] A transient path planning module is connected to the state perception module and the optimized path database respectively. It is used to respond to the target load and response time limit input by the external input, and to match and call the optimal thermodynamic state path and the time-series feedforward control sequence from the optimized path database based on the actual thermodynamic state vector.

[0013] A path tracking execution module, which is connected to the transient path planning module, the state perception module, and the physical execution layer, is used to receive the optimal thermodynamic state path and the timing feedforward control sequence, and generate a correction control signal based on the state deviation vector between the actual thermodynamic state vector and the optimal thermodynamic state path. The timing feedforward control sequence is then superimposed with the correction control signal to generate the final timing control command, which is then sent to the physical execution layer.

[0014] Preferably, the state-aware module includes:

[0015] A set of sensors includes at least an intake pressure sensor, an exhaust pressure sensor, an intake temperature sensor, and a condenser outlet liquid temperature sensor;

[0016] The data processing unit receives the measured values ​​from the sensor and calculates the suction superheat and condenser outlet subcooling through property calculations, so as to use the suction pressure, discharge pressure, suction superheat, and condenser outlet subcooling as components of the actual thermodynamic state vector.

[0017] Preferably, the optimized path database includes a conventional optimal path database, which is used to store optimal transfer data with the starting state region, the target state region and the response time limit as multi-dimensional index keys;

[0018] The response time limit index is used to pre-balance the regulation speed and regulation efficiency, so as to provide the optimal thermodynamic state path with different transient costs for the same initial state region and target state region.

[0019] Preferably, the optimized path database further includes a special operating condition path database, which stores at least the defrost entry path, the defrost exit path, and the timing feedforward control sequence corresponding to the path.

[0020] Preferably, the transient path planning module is further used for:

[0021] Monitor one or more special operating condition triggers;

[0022] When the special operating condition trigger is activated, the target load and the response time limit are ignored, and a preset path is forcibly called from the special operating condition path library as the optimal thermodynamic state path.

[0023] Preferably, the transient path planning module is further configured to, when there is no exact matching path in the optimized path database:

[0024] Determine the time penalty factor based on the response time limit;

[0025] Based on the time penalty factor and the basic transfer cost pre-stored in the optimized path database, the dynamic modulation cost is calculated and generated online.

[0026] A real-time pathfinding algorithm is executed based on the dynamic modulation cost to calculate and generate the optimal thermodynamic state path and the time-series feedforward control sequence.

[0027] Preferably, the path tracing execution module is used for:

[0028] Calculate the first derivative of the state deviation vector, which reflects the deviation trend;

[0029] Based on the state deviation vector and its first derivative, the correction control signal is calculated and generated using a pre-calibrated inertial correction compensator.

[0030] Preferably, the path tracking execution module generates the correction control signal through a multivariable PID controller, which calculates the signal based on the proportional, integral, and derivative terms of the state deviation vector.

[0031] Preferably, the physical execution layer includes:

[0032] Variable frequency compressor, electronic expansion valve, and at least an outdoor fan and an indoor fan;

[0033] The final timing control command is a timing vector, which includes at least a compressor frequency command for the variable frequency compressor, an electronic expansion valve opening command for the electronic expansion valve, and a fan speed command for the fan.

[0034] A second aspect of the present invention provides an adaptive load-adjusting air-source heat pump control method, using the system described in any of the preceding claims, the method comprising the following steps:

[0035] Real-time acquisition of physical parameters and calculation of actual thermodynamic state vectors;

[0036] In response to the target load and response time limit of the external input, and based on the actual thermodynamic state vector, the optimal thermodynamic state path and the timing feedforward control sequence corresponding to the optimal thermodynamic state path are matched and called from the optimization path database. The optimization path database pre-stores the optimal thermodynamic state path and the corresponding timing feedforward control sequence.

[0037] Based on the state deviation vector between the actual thermodynamic state vector and the optimal thermodynamic state path, a correction control signal is generated.

[0038] The timing feedforward control sequence is superimposed with the correction control signal to generate the final timing control command, and the physical execution layer is controlled to execute a thermodynamic cycle according to the final timing control command.

[0039] This invention provides an adaptive load-regulating air-source heat pump system and its control method. It offers the following advantages:

[0040] 1. This invention uses a response time limit as a key index, enabling the transient path planning module to match optimal thermodynamic state paths with different costs for the same adjustment task within the optimized path database. This allows the system to achieve an adaptive balance between adjustment speed and operating energy efficiency. When user demand is urgent, the system automatically calls the time-optimal path; when demand is moderate, it automatically calls the energy-optimal path.

[0041] 2. This invention completes the highly complex multi-objective optimization calculation offline and stores the optimal solution set in an optimized path database. During online operation, the transient path planning module mainly performs high-speed table lookups and matching, rather than real-time complex pathfinding calculations. This reduces the online computing burden on the controller, making this invention easy to deploy and implement on cost-sensitive embedded systems.

[0042] 3. This invention achieves high-precision and high-robustness path tracking through a path tracking execution module. This module superimposes the time-series feedforward control sequence from the database with the correction control signal calculated in real time based on the state deviation vector, forming a composite control. Feedforward control ensures rapid response to the target path and baseline tracking, while closed-loop correction overcomes the effects of actual system model mismatch and external environmental disturbances, ensuring that the system always operates stably along the optimal path.

[0043] 4. This invention improves the system's operational reliability under severe disturbances such as defrosting by setting up a special operating condition path library in the optimized path database. For specific processes such as defrosting entry and exit, the database pre-stores dedicated, optimized transition paths, avoiding control oscillations or inefficiencies in conventional load regulation logic when dealing with such highly nonlinear processes, thus ensuring the system's stability and economy during special operating condition switching. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the system architecture of the present invention;

[0045] Figure 2 This is a schematic diagram of the physical execution layer structure of the present invention;

[0046] Figure 3 This is a schematic diagram showing the connection between the state perception module and the physical execution layer of the present invention;

[0047] Figure 4 This is a functional logic flowchart of the transient path planning module of the present invention;

[0048] Figure 5 This is a functional block diagram of the path tracing execution module of the present invention. Detailed Implementation

[0049] The technical solutions in 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.

[0050] Please see the appendix Figure 1 , Figure 1 This is a schematic diagram of a system architecture according to an embodiment of the present invention. The present invention provides an adaptive load-adjusting air source heat pump system, which may include: a physical execution layer; a state perception module; an optimized path database; a transient path planning module; and a path tracking execution module.

[0051] This system defines the dynamic load regulation process of the heat pump as a path optimization problem in the multidimensional thermodynamic state space (TSS).

[0052] This system optimizes the path database by pre-storing the optimal transient path, and achieves online optimal path matching and tracking through the collaboration of the transient path planning module and the path tracking execution module.

[0053] The optimized path database is pre-built through offline modeling. It stores TSS manifold data that characterizes the healthy operation of the system, as well as the core conventional optimal path library and special operating condition path library.

[0054] The transient path planning module is responsible for online decision-making. It receives the current state from the state awareness module and the external target load and response time limit.

[0055] The path tracing execution module is responsible for receiving the planned path and sending timing control instructions to the physical execution layer.

[0056] The present invention also provides an adaptive load adjustment air source heat pump control method, which generally includes an offline modeling stage and an online execution stage.

[0057] Offline modeling phase. This phase is completed during the R&D or calibration process and is used to build an optimized path database.

[0058] This stage first defines the system's multidimensional thermodynamic state vector, which consists of key thermodynamic parameters such as compressor suction pressure, discharge pressure, suction superheat, and condenser outlet subcooling.

[0059] Next, a thermodynamic state-space (TSS) manifold is constructed, which is the set of all healthy, quasi-steady-state operating points of the system in the state space.

[0060] Subsequently, a basic transfer cost matrix is ​​defined. This basic transfer cost matrix represents the minimum transient energy cost consumed by the system to transition from one state to another. This cost can be obtained by integrating the instantaneous power consumption of the compressor during the transition.

[0061] Finally, based on the TSS manifold and the basic transfer cost matrix, a large number of optimal transient paths are pre-generated through offline optimization calculations and stored in the conventional optimal path library and the special working condition path library.

[0062] Online execution phase. This online execution phase is jointly executed by the transient path planning module and the path tracing execution module during device runtime.

[0063] State awareness and target interpretation. The state awareness module acquires the current thermodynamic state in real time. The transient path planning module receives the external target load and response time limit, and interprets them into the target state region on the TSS manifold.

[0064] Path matching and invocation. The transient path planning module first checks for the presence of special operating conditions (such as defrosting signals). If triggered, it retrieves the preset optimal path and its corresponding timing control sequence from the special operating condition path library.

[0065] For routine adjustments, the transient path planning module uses the current state, target state region, and response time limit as query indexes to quickly match and call an optimal transient path and its corresponding timing control sequence in the conventional optimal path library.

[0066] The response time limit serves as a key index, ensuring that the matched path is the optimal solution that satisfies the time constraint. For example, with the same start and end points, a shorter response time limit will match a time-priority path, while a longer response time limit will match an energy-efficiency-priority path.

[0067] In an optional embodiment, if no exact match is found in the conventional optimal path library, the transient path planning module can dynamically modulate the basic transfer cost based on the response time limit to perform small-scale real-time interpolation or local pathfinding. This dynamically modulated cost is determined by both the basic energy cost and a time penalty factor, where the time penalty factor is determined by the response time limit.

[0068] Path tracking and correction. The path tracking execution module receives the target path and the feedforward control sequence. This module calculates the state deviation vector between the actual state and the target path in real time. Based on this state deviation vector, the path tracking execution module generates a correction control signal.

[0069] In a preferred embodiment, the correction control signal may employ thermodynamic inertial feedforward compensation, which is calculated based on the state deviation vector and its rate of change. In another embodiment, robust feedback correction (e.g., PID) may be employed.

[0070] The final instruction sent to the physical execution layer is a superposition of feedforward and correction instructions. This mechanism ensures that when the system is subjected to external disturbances or model mismatch, its actual operating state is still actively corrected to run along the target path.

[0071] Please see the appendix Figure 2 , Figure 2 This is a schematic diagram of the physical execution layer structure according to an embodiment of the present invention.

[0072] The physical execution layer constitutes the controlled object of the system of this invention, namely the hardware body of the air source heat pump. This physical execution layer is responsible for executing the thermodynamic cycle in response to control commands.

[0073] In one embodiment, the physical execution layer includes at least: a variable frequency compressor, an outdoor heat exchanger, an indoor heat exchanger, an electronic expansion valve (EEV), an outdoor fan, an indoor fan, and a four-way reversing valve. These components are interconnected via refrigerant piping to form a complete, switchable refrigerant circuit.

[0074] The key function of the physical execution layer lies in its inclusion of multiple controllable actuators that receive final timing control instructions from the path tracing execution module. This sequence of instructions is then translated into physical hardware actions.

[0075] Final timing control instructions It is a timing vector containing multiple control channels. This vector must include at least... compressor frequency command at any time , Electronic expansion valve opening command at any time ,as well as outdoor fan speed command at any time Indoor fan speed command .

[0076] Specifically, the variable frequency compressor is based on The electronic expansion valve operates to regulate the refrigerant circulation rate and pressure rise. Adjust the opening degree to control the refrigerant throttling and flow distribution. The outdoor and indoor fans operate according to their respective speed commands. and It operates to control the airflow through the outdoor and indoor heat exchangers.

[0077] The four-way reversing valve also receives switching commands from the path tracking execution module to switch between heating and cooling modes, or to execute defrost cycle paths (such as those called from the special operating condition path library) from the path library. or ).

[0078] The physical execution layer through... The timing coordination of commands from each component ensures a controlled and continuous change in the state of the internal working fluid (refrigerant). This change is reflected in the actual thermodynamic state measured by the state-sensing module. Therefore, the physical execution layer constitutes the method of the present invention, enabling... Follow the target path The physical basis of movement.

[0079] Please see the appendix Figure 3 , Figure 3This is a schematic diagram of the connection between the state awareness module and the physical execution layer according to an embodiment of the present invention.

[0080] The state awareness module, deployed above the physical execution layer, consists of a set of sensors and a data processing unit. Its function is to acquire key physical parameters of the physical execution layer in real time during operation, and through calculation, generate and output a real thermodynamic state vector characterizing the current operating state of the system. .

[0081] Should The vector is provided to the transient path planning module for path decision-making and to the path tracking execution module for deviation monitoring. The vector data structure and the state vector defined in the offline modeling phase Maintain consistency.

[0082] In this invention, the specific implementation of the state-aware module may include the following steps:

[0083] Physical parameter acquisition. In a specific embodiment, the state awareness module includes at least the following sensors, deployed at the corresponding pipeline nodes of the physical execution layer:

[0084] Suction pressure sensor: Installed in the compressor's suction line, used to measure... Inhalation pressure at any time .

[0085] Exhaust pressure sensor: Installed in the compressor's exhaust line, used to measure... exhaust pressure at any time .

[0086] Suction temperature sensor: Installed in the compressor's suction line, its measuring point is adjacent to the suction pressure sensor, used to measure... Inhalation temperature at any time .

[0087] Condenser outlet temperature sensor: Installed on the liquid phase outlet pipe of the condenser (the heat exchanger corresponding to heating or cooling mode) in the physical actuator layer, used to measure... Condenser outlet temperature at time .

[0088] State vector calculation. The data processing unit within the state perception module is responsible for receiving the raw sensor data acquired during physical parameter acquisition. and calculate and generate in real time vector.

[0089] In vectors and The components can be directly measured using the values ​​from the intake pressure sensor and the exhaust pressure sensor.

[0090] The intake superheat in the vector The component is not measured directly, but calculated by the data processing unit using the following formula:

[0091] ;

[0092] In the formula, for The constant temperature of the inhaled air; for Inhalation temperature at any given moment; for Inhalation pressure at any given moment; It is a function that represents the input inspiratory pressure. The calculated corresponding saturation temperature; This is a time-time variable. The value is calculated in real time by the data processing unit by looking up the pre-stored refrigerant property table or calling the refrigerant property state equation (such as the RK equation or the Helmholtz equation).

[0093] Condenser outlet subcooling in the vector The component is also calculated by the data processing unit using the following formula:

[0094] ;

[0095] In the formula, for The degree of subcooling at the condenser outlet at any given time; for Discharge pressure (or condensation pressure) at any given time; It is a function that represents the output based on the input exhaust pressure. The calculated corresponding saturation temperature; for The condenser outlet temperature at any given time; For time variables.

[0096] State vector output. The state-aware module outputs the complete vector calculated during the state vector computation:

[0097] ;

[0098] In the formula, for The actual thermodynamic state vector at time 1; for Inhalation pressure at any given moment; for The exhaust pressure at any given moment; for The constant temperature of the inhaled air; for The degree of subcooling at the condenser outlet at any given time; For time variables; This represents a column vector that contains components and may contain other state components.

[0099] The data is output to the transient path planning module and the path tracking execution module in real time and continuously at a predetermined sampling period.

[0100] The optimized path database is the core knowledge base of the system of this invention, physically stored in the controller's non-volatile memory (e.g., Flash or EEPROM). This path database is not a simple parameter record, but a multi-dimensional lookup table generated through offline precise modeling and massive optimization calculations, encompassing the optimal transient adjustment strategies for the system under various operating conditions. It provides the online transient path planning module with optimal solutions that can be directly accessed without requiring complex real-time pathfinding calculations.

[0101] The optimized path database is logically divided into at least two core sub-databases: the conventional optimal path database and the special working condition path database.

[0102] The optimal path library stores the optimal transfer strategies between different load demands under normal heating or cooling modes. This optimal path library employs a multi-dimensional index structure, and its query index key includes at least three dimensions: the initial state region... Target state area and response time limit .

[0103] Among them, the initial state region and target state region For continuous thermodynamic space-state (TSS) manifolds Region identification after discretization and mesh generation. Response time limit. It is a discrete classification of the urgency of external load regulation.

[0104] For each specific set of index keys The standard optimal path library stores a unique set of optimal transfer data values. These optimal transfer data values ​​contain two paired time series sequences:

[0105] One is the optimal thermodynamic state path. It describes the system from Transferred to And satisfy Under constraints, the theoretically lowest state change trajectory with the lowest total transient transition cost (including energy and time costs) is:

[0106] ;

[0107] In the formula, The optimal thermodynamic state path represents a temporal sequence of state vectors; It is a time variable; for The optimal thermodynamic state vector at time t; Let be a series of discrete time points on the optimal path, where This represents the starting time of the path. This represents the end time of the path.

[0108] The second is the corresponding timing feedforward control sequence. It describes the state path to achieve the above. The combination of timing actions that each executor in the physical execution layer needs to execute:

[0109] ;

[0110] Here, This is a timing feedforward control sequence, representing a timing control vector sequence; It is a time variable; for The feedforward control vector at any given time contains, for example, specific instruction values ​​of the actuator such as compressor frequency and electronic expansion valve opening; Let be the discrete time points on the optimal path, where This represents the starting time of the path. This represents the end time of the path.

[0111] By introducing response time limits in the index The conventional optimal path library implements a pre-balancing of regulation speed and regulation energy efficiency. For the same... and The database will store multiple alternative paths: when When the size is small, it stores high-energy-consuming but fast paths; when... When the size is large, it stores low-energy but slower paths.

[0112] The special operating condition path library is specifically designed to store strategies for specific transient processes that are not targeted at regular loads. This library contains at least the defrost ingress path. and defrost exit path And their corresponding timing feedforward control sequences. These paths are specifically optimized for the thermodynamic disturbance process of defrosting, aiming to complete the switching between heating and defrosting modes with the lowest energy consumption.

[0113] In a preferred embodiment, the optimized path database may also selectively store TSS manifold data. and basic transfer cost matrix This basic data can be used by the transient path planning module to perform path planning when no exact matching path is found. Perform small-scale real-time interpolation calculations to generate an approximate optimal path.

[0114] Please see the appendix Figure 5 , Figure 5 This is a functional logic flowchart of a transient path planning module according to an embodiment of the present invention.

[0115] The transient path planning module is the core decision-making component of the system during online operation. As a logical functional unit, this module is responsible for receiving the actual thermodynamic state vector from the state sensing module. And target loads from external sources (such as user interfaces or upper-level building management systems). and response time .

[0116] The core function of the transient path planning module is to perform high-speed queries, matching, or calculations in the optimized path database based on the aforementioned real-time input, and finally output a set of optimal adjustment instructions to the path tracking execution module. These adjustment instructions specifically include the optimal thermodynamic state path. and its paired timing feedforward control sequence .

[0117] In one specific embodiment, the internal execution logic of the transient path planning module may include the following steps:

[0118] Target and state interpretation. This step converts the continuous input signal into discrete indices recognizable by the optimized path database. The transient path planning module receives... and map it to the TSS manifold The initial state region on Meanwhile, the transient path planning module receives... And interpret it as a TSS manifold The target state region on The transient path planning module also receives... It is then normalized or discretized into the standard time levels defined in the database.

[0119] Operating condition management. This step handles the arbitration between special operating conditions and routine load adjustments. The transient path planning module monitors one or more special operating condition triggers in real time. (For example, a defrost detection signal from a sensor). If Once activated, the transient path planning module will immediately interrupt regular path matching and invocation, and instead execute special working condition path invocation.

[0120] Regular path matching and invocation. This step is performed by the transient path planning module when no special operating condition trigger is detected in the operating condition management. This transient path planning module uses the results interpreted in the target and state interpretation... It serves as a three-dimensional query index, enabling high-speed lookups within the conventional optimal path library of the optimized path database.

[0121] Once a matching entry is successfully found, the transient path planning module will read from that entry. and Two time sequences are generated and sent as output to the path tracing execution module.

[0122] Special operating condition path invocation. This occurs during operating condition status management. When activated, the transient path planning module performs this step. This transient path planning module will ignore... and and according to The type (e.g., defrost entry or defrost exit) is forcibly invoked from the special operating condition path library of the optimized path database, using a preset path. and its corresponding Sequence, and its (as) and Send to the path tracing execution module.

[0123] In a preferred embodiment, the transient path planning module also possesses path interpolation and generation capabilities to address situations where an exact matching index cannot be found during regular path matching and invocation. In this case, the transient path planning module can search for the nearest few path indices in the optimized path database and perform [further processing / analysis]. and The sequence is weighted and interpolated to generate an approximate optimal solution.

[0124] In another, more advanced embodiment, the transient path planning module can perform a small range of real-time pathfinding calculations. In this case, the transient path planning module can call upon data stored in a database. and Basic data, and based on the current Dynamic cost modulation is performed, and its modulation cost is... It can be defined online by the following formula:

[0125] ;

[0126] In the formula, The dynamic modulation cost is the final composite cost that the transient path planning module uses when executing the real-time pathfinding algorithm. The basic transient energy cost refers to the energy cost of a system transitioning from its thermodynamic state. Transition to thermodynamic state Basic energy consumption; The fundamental transition time refers to the time from the thermodynamic state of the system. Transition to thermodynamic state Required base time; The response time limit for external inputs is used to characterize the urgency of load regulation; The time penalty factor is the response time limit. A function used for online trade-offs between energy and time costs, including a time penalty factor. With response time limit It increases as it shortens; The initial thermodynamic state; The target thermodynamic state.

[0127] The transient path planning module then based on As edge weights, within a local region of the TSS manifold (e.g.) and Between), execute a real-time pathfinding algorithm (such as A* algorithm or Dijkstra's algorithm) to compute the new... And simultaneously solve This embodiment offers greater control flexibility, but places higher demands on the controller's online computing capabilities.

[0128] The path tracking execution module is the core of the online execution and closed-loop correction of the system of this invention. The function of this path tracking execution module is to receive the optimal adjustment instructions (i.e., the optimal thermodynamic state path and the timing feedforward control sequence) from the transient path planning module, and combine them with the actual state from the state perception module to generate and send the final timing control instructions to the physical execution layer.

[0129] This path tracing execution module ensures that the actual thermodynamic state of the physical execution layer accurately follows the predetermined operation. The core problem solved by this module is to overcome the model mismatch between the offline model of the optimized path database and the actual physical characteristics of the physical execution layer, and to compensate for external unmodeled disturbances encountered during operation, thereby achieving robust tracking of the target path.

[0130] In one specific embodiment, the internal execution logic of the path tracing execution module may include the following steps:

[0131] The path tracing execution module receives the target path and actual state in real time from the transient path planning module. Simultaneously, it receives the actual thermodynamic state vector at the current moment from the state awareness module.

[0132] Calculate the state deviation vector. This path tracing execution module calculates the state deviation vector in real time based on the input of the received target path and the actual state.

[0133] ;

[0134] In the formula, for The state deviation vector at any given time represents the instantaneous deviation between the actual state and the target path in each thermodynamic dimension; for The target state point on the optimal thermodynamic state path at any given time is the ideal state value output by the transient path planning module; for The actual thermodynamic state vector at any given moment is calculated and fed back in real time by the state-aware module.

[0135] Generate correction control signals. This path tracking execution module calculates the correct control signals based on the calculated state deviation vector. Generate a correction control signal This correction control signal is used for compensation. The deviations reflected.

[0136] In one embodiment, This can be generated by a multivariable feedback controller (such as a PID controller well-known in the art). The multivariable feedback controller is based on the state deviation vector. The correction signal is calculated using proportional, integral, and derivative terms, and the deviation is eliminated through hysteresis feedback.

[0137] In another preferred embodiment, to overcome the large inertia and hysteresis characteristics of processes such as refrigerant migration and pressure build-up in the physical execution layer, This is achieved using a thermodynamic inertial feedforward correction. The aim is to anticipate deviations caused by system inertia and compensate for them in advance.

[0138] ;

[0139] In the formula, for The correction control signal generated at each moment is a control vector used to compensate for state deviations; for The thermodynamic inertial feedforward correction signal at time t is A preferred embodiment; An inertial correction compensator is typically a pre-calibrated nonlinear function or model whose inputs are state deviations and their rates of change, and whose output is the desired feedforward correction. State deviation vector Regarding time The first derivative, i.e., the rate of change of state deviation, reflects the trend of deviation. This model can be based on the system's thermodynamic time constant (e.g., the mass migration time constant). Pressure response time constant )Establish.

[0140] The final command is then superimposed and issued. The path tracking execution module superimposes the received target path and the feedforward control sequence received in the actual state with the correction control signal generated in the correction control signal generation process to synthesize the final physical execution command.

[0141] The physical execution instruction is a complete timing control vector, which is then sent to each actuator in the physical execution layer, thus forming a composite control structure of feedforward + feedback or feedforward + feedforward correction.

[0142] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An air source heat pump system with adaptive load regulation, characterized by, The system comprises: a physical execution layer for executing a thermodynamic cycle; a state sensing module connected with the physical execution layer for collecting physical parameters in real time and calculating an actual thermodynamic state vector; an optimal path database for storing a pre-generated optimal thermodynamic state path and a time-sequential feedforward control sequence corresponding to the optimal thermodynamic state path; a transient path planning module connected with the state sensing module and the optimal path database respectively for responding to an externally input target load and response time limit and matching and calling the optimal thermodynamic state path and the time-sequential feedforward control sequence from the optimal path database based on the actual thermodynamic state vector; a path tracking execution module connected with the transient path planning module, the state sensing module and the physical execution layer respectively for receiving the optimal thermodynamic state path and the time-sequential feedforward control sequence, generating a correction control signal based on a state deviation vector between the actual thermodynamic state vector and the optimal thermodynamic state path, and superimposing the time-sequential feedforward control sequence and the correction control signal to generate a final time-sequential control instruction and issuing the final time-sequential control instruction to the physical execution layer; the optimal path database comprises a regular optimal path library for storing optimal transition data indexed by a starting state region, a target state region and the response time limit as multi-dimensional index keys; the response time limit is used to realize a prior trade-off between adjustment speed and adjustment energy efficiency to provide the optimal thermodynamic state path with different transient costs for the same starting state region and target state region; the optimal path database further comprises a special working condition path library for storing at least a defrosting entry path, a defrosting exit path and a time-sequential feedforward control sequence corresponding to the defrosting entry path and the defrosting exit path.

2. An air source heat pump system with adaptive load regulation according to claim 1, wherein, The state sensing module comprises: a set of sensors including at least a suction pressure sensor, a discharge pressure sensor, a suction temperature sensor and a condenser outlet temperature sensor; a data processing unit for receiving measurement values of the sensors and obtaining suction superheat and condenser outlet subcooling through property calculation to take the suction pressure, the discharge pressure, the suction superheat and the condenser outlet subcooling as components of the actual thermodynamic state vector.

3. An air source heat pump system with adaptive load regulation according to claim 1, wherein, The transient path planning module is further used for: monitoring one or more special working condition triggers; when the special working condition triggers are activated, ignoring the target load and the response time limit and forcibly calling a preset path from the special working condition path library as the optimal thermodynamic state path.

4. An air source heat pump system with adaptive load regulation according to claim 1, wherein, The transient path planning module is further used for, when there is no accurate matching path in the optimal path database: determining a time penalty factor based on the response time limit; calculating a dynamic modulation cost on line according to the time penalty factor and a basic transition cost pre-stored in the optimal path database; performing a real-time path tracking algorithm based on the dynamic modulation cost to calculate the optimal thermodynamic state path and the time-sequential feedforward control sequence.

5. An air source heat pump system with adaptive load regulation according to claim 1, wherein, The path tracking execution module is configured to: calculate a first derivative of the state deviation vector, the first derivative reflecting a deviation trend; based on the state deviation vector and the first derivative of the state deviation vector, generate the correction control signal through a pre-calibrated inertial correction compensator.

6. An air source heat pump system with adaptive load regulation according to claim 1, wherein, The path tracking execution module generates the correction control signal through a multivariable PID controller, which is calculated based on proportional, integral and derivative terms of the state deviation vector.

7. An air source heat pump system with adaptive load regulation according to claim 1, wherein, The physical execution layer comprises: a variable frequency compressor, an electronic expansion valve, and at least an outdoor fan and an indoor fan; The final time-sequential control instruction is a time-sequential vector, which comprises a compressor frequency instruction for the variable frequency compressor, an electronic expansion valve opening degree instruction for the electronic expansion valve, and a fan rotating speed instruction for the outdoor fan and the indoor fan.

8. An air source heat pump control method with adaptive load regulation, characterized by, The method comprises the following steps: real-time acquisition of physical parameters and calculation of an actual thermodynamic state vector; in response to an externally input target load and response time limit, and based on the actual thermodynamic state vector, matching and calling an optimal thermodynamic state path and a time-sequential feedforward control sequence corresponding to the optimal thermodynamic state path from an optimization path database, the optimization path database pre-storing the optimal thermodynamic state path and the corresponding time-sequential feedforward control sequence; based on a state deviation vector between the actual thermodynamic state vector and the optimal thermodynamic state path, generating a correction control signal; superimposing the time-sequential feedforward control sequence and the correction control signal to generate a final time-sequential control instruction, and controlling a physical execution layer to execute a thermodynamic cycle according to the final time-sequential control instruction.

Images