Thermostatic controllable load regulation method and system based on multi-agent control

Abstract

The application discloses a kind of constant temperature controllable load regulation and control method and system based on multi-agent control, comprising: the performance parameter and refrigeration performance index of each constant temperature controllable load in large-scale constant temperature controllable load cluster are collected, and the thermodynamic model corresponding to the room where each constant temperature controllable load is located is constructed, the large-scale constant temperature controllable load cluster is defined as multi-agent system, and control framework and communication model for controlling multi-agent system through aggregator are established, nonlinear consistency controller is constructed according to preset constant temperature controllable load state variable calculation method, nonlinear saturation consistency algorithm, relay node state update model and fallback mechanism, the initial state of the large-scale constant temperature controllable load cluster is obtained, according to the initial state and the thermodynamic model, the nonlinear consistency controller is regulated and controlled, for realizing the unified regulation of large-scale constant temperature controllable load.

Description

Technical Field

[0001] This invention relates to the field of microgrid technology, and more specifically, to a method and system for temperature-controlled load regulation based on multi-agent control. Background Technology

[0002] A microgrid is a cluster of distributed resources, namely renewable energy sources (RESs), energy storage systems (ESSs), and distribution-level controllable loads. Depending on its size, it can comprise a complete distribution network or a building with distributed resources. When operating in grid-connected mode, a microgrid can be considered a single entity. A group of interconnected microgrids can form a microgrid community, or gridded microgrid, collectively providing ancillary services or operational support to the main grid, such as frequency regulation and emergency demand response.

[0003] A major challenge facing microgrids is the uncertain power variation caused by renewable energy sources (RESs) and load demand. This makes the total power injection / consumption from the microgrid to the main grid a stochastic and uncontrollable variable, hindering the microgrid's participation in power system ancillary services or operational support. Controlling the microgrid's participation in power system demand response can be approached in two ways: one relies on energy storage systems, which can mitigate power fluctuations, but their capacity and widespread application remain limited due to high investment costs. The other approach involves deploying real-time control of loads at the customer end for demand response.

[0004] However, for large-scale spatially distributed systems, centralized control architectures have a large computational and communication burden, making them more suitable for operation scheduling rather than real-time control. In addition, centralized control systems are inherently susceptible to single points of failure and communication failures.

[0005] Driven by intelligent sensing, communication, and computing, distributed control of network systems has received widespread attention in fields such as smart buildings and smart grids. Due to the complexity of modern building structures and the large number of TCLs (Transmission Control Units), there is an urgent need for highly scalable and plug-and-play distributed control solutions. Furthermore, since there may be no central controller in a community, distributed control is also suitable for coordinating between various buildings. Summary of the Invention

[0006] To address the aforementioned issues, this invention discloses a method and system for regulating a constant-temperature controllable load based on multi-agent control, which enables unified regulation of large-scale constant-temperature controllable loads.

[0007] To achieve the above objectives, this invention discloses a method for controlling a constant temperature load based on multi-agent control, comprising:

[0008] The performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster are collected, and the room where each constant temperature controllable load is located is thermodynamically modeled according to the preset equivalent thermal parameter model, so as to construct the thermodynamic model corresponding to each room.

[0009] The large-scale constant temperature controllable load cluster is defined as a multi-agent system, and a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system are established.

[0010] A nonlinear consistency controller is constructed based on the preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consistency algorithm, a relay node state update model, and a backoff mechanism.

[0011] The initial state of the large-scale isothermal controllable load cluster is obtained, and each isothermal controllable load is regulated by the nonlinear consistency controller based on the initial state and the thermodynamic model.

[0012] This invention discloses a method for regulating a constant-temperature controllable load based on multi-agent control. First, the performance parameters and cooling performance indicators of each constant-temperature controllable load in a large-scale constant-temperature controllable load cluster are obtained. Then, a thermodynamic model corresponding to the room where the load is located is constructed according to a preset equivalent thermal parameter model. By performing thermodynamic modeling on the room where the load is located, the case of heterogeneous constant-temperature controllable loads is considered, which improves the practicality of the regulation method.

[0013] Simultaneously, during the regulation process, each load is treated as an intelligent agent. A control framework for controlling the multi-agent system through an aggregator and a communication model for communication between agents are constructed, realizing unified regulation of large-scale constant-temperature controllable loads. Based on a preset constant-temperature controllable load state variable calculation method, a nonlinear saturation consensus algorithm, a relay node state update model, and a backoff mechanism, a nonlinear consensus controller is constructed. This enables unified regulation of large-scale constant-temperature controllable loads under nonlinear saturation factors when regulating the load through the nonlinear consensus controller, improving the engineering requirements capability of the method. At the same time, the backoff mechanism ensures that user experience is considered during the regulation process, improving the user experience.

[0014] As a preferred example, the performance parameters and cooling performance indicators of each constant-temperature controllable load in the large-scale constant-temperature controllable load cluster include:

[0015] The cooling performance index of each constant-temperature controllable load is calculated using a preset cooling performance index function; wherein the expression of the cooling performance index function is:

[0016]

[0017] Wherein, the γ i (t) represents the refrigeration performance index of the i-th constant-temperature controllable load, where τ i (t) represents the operating efficiency of the i-th isothermal controllable load, where ω i (t) represents the operating frequency of the i-th constant-temperature controllable load, where a i1 ,β i1 Let a be the operating power coefficient of the i-th isothermal controllable load. i2 ,β i2 Let be the coefficient of performance for the i-th constant-temperature controllable load;

[0018] The performance parameters of each constant-temperature controllable load are calculated using a preset performance parameter function; wherein, the expression of the performance parameter function is:

[0019] H i (t)=γ i (t) / τ i (t)

[0020] Wherein, the H i (t) represents the performance parameter.

[0021] This invention calculates the performance parameters and refrigeration performance indicators to enable the subsequent construction of the thermodynamic model, so that the actual operating conditions of the load can be taken into account when regulating the load, thereby improving the regulation effect.

[0022] As a preferred example, the thermodynamic modeling of each room containing a constant-temperature controllable load is performed based on a preset equivalent thermal parameter model to construct a thermodynamic model corresponding to each room, including:

[0023] The thermodynamic model is constructed using a pre-defined equivalent thermal parameter model; wherein the expression of the thermodynamic model is:

[0024]

[0025] Where i∈π represents the set of isothermal controllable loads, t∈θ represents the set of time points, and σ i ρ represents the equivalent heat capacity of the room containing the i-th temperature-controlled load. i δ represents the equivalent thermal resistance of the room containing the i-th temperature-controlled load. i δ(t) represents the indoor temperature of the room containing the i-th constant temperature controllable load at time t, δ0(t) represents the outdoor temperature at time t, and γ i This represents the refrigeration performance index of the i-th constant temperature controllable load.

[0026] The present invention constructs the thermodynamic model to take into account heterogeneous isothermal controllable loads located in different rooms, so that the control method provided by the present invention can uniformly control the heterogeneous isothermal controllable loads, improve the practicality of control, and adapt to different control scenarios.

[0027] As a preferred example, the establishment of the control framework for controlling the multi-agent system via an aggregator and the communication model between pairs of agents in the multi-agent system includes:

[0028] The control framework is constructed based on the composition, communication, and room temperature of the temperature-controlled loads in the multi-agent system. The composition involves a temperature-controlled load cluster composed of heterogeneous temperature-controlled loads. The communication involves nonlinear saturation factors in information transmission when two temperature-controlled load agents communicate. The room temperature involves the gradual increase in indoor temperature beyond the preset comfort value when the load adjustment time or performance is too high.

[0029] Several relay nodes are set up in the multi-agent system, and the relay nodes are defined as communication hubs; wherein, the relay nodes communicate bidirectionally with several temperature-controlled loads in the large-scale temperature-controlled load cluster, with the temperature-controlled loads and other adjacent temperature-controlled loads, with each pair of relay nodes, and with the aggregator.

[0030] The aggregator sends the preset overall load control target to the relay node connected to the aggregator, and the relay node sends the overall load control target to each agent in the multi-agent system, so that each agent can reach a consensus on the overall load control target through bidirectional communication.

[0031] This invention constructs the control framework based on the actual load composition, communication conditions, and room conditions, making the load adjustment more realistic and improving the load control effect. At the same time, it constructs a two-way communication network between intelligent agents to enable consistent communication for load control and improve the efficiency of control.

[0032] As a preferred example, the construction of a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consensus algorithm, a relay node state update model, and a backoff mechanism includes:

[0033] The state variables of each temperature-controlled load are calculated and determined based on its operating frequency information, physical constraints, application scenario requirements, and control mode. The state variables of each temperature-controlled load have different calculation methods under both upward and downward adjustment modes. The expression for the calculation method of the temperature-controlled load state variables is as follows:

[0034]

[0035] Wherein, the x i ∈x=[x1,x2,...,x n ], where n is the total number of isothermal controllable load intelligent agents, and x i (t)∈[0,1], representing the state information of the i-th agent, the It is the initial operating frequency of the i-th temperature-controlled load compressor. and These represent the lower limit and upper limit of the inherent operating frequency of the compressor with the i-th constant temperature controllable load, respectively.

[0036] The present invention can normalize the regulation capability of heterogeneous constant temperature controllable loads based on the state variable calculation method of the aforementioned constant temperature controllable load, and use a multi-agent cooperative control method to control the constant temperature controllable load based on this, thereby realizing the connection of heterogeneous constant temperature controllable loads using the same control strategy.

[0037] As a preferred example, the construction of a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consensus algorithm, a relay node state update model, and a backoff mechanism includes:

[0038] The nonlinear saturated consensus algorithm is constructed based on a preset standard average consensus algorithm and a nonlinear saturation function; wherein, the expression of the standard average consensus algorithm is:

[0039]

[0040] Wherein, the Ω i It is the set of neighboring nodes of the i-th isothermal controllable load intelligent agent, wherein a ij The x is an element of the adjacency matrix A of the multi-agent communication network topology. i and x j These are the state of the i-th constant temperature controllable load agent and the state information of the j-th neighbor node of the i-th constant temperature controllable load agent, respectively.

[0041] The expression for the nonlinear saturation function is:

[0042]

[0043] in, It is a nonlinear saturation function, the y i (t) is an intermediate variable, the aforementioned n is the total number of agents;

[0044] The expression for the nonlinear saturation consensus algorithm is:

[0045]

[0046]

[0047] Wherein, ξ is the control gain of the relay node, and y i (t) is an intermediate variable;

[0048] Based on the standard average consensus algorithm and the nonlinear saturated consensus algorithm, a matrix of consensus algorithms with nonlinear protocols is obtained; wherein, the expression of the matrix is:

[0049]

[0050] in,

[0051] This invention combines nonlinear saturation with a standard average consensus algorithm to construct a nonlinear saturation consensus algorithm, thereby enabling unified adjustment of large-scale isothermal controllable loads under nonlinear saturation factors, and giving the method of this invention a stronger ability to adapt to engineering needs.

[0052] As a preferred example, the construction of a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consensus algorithm, a relay node state update model, and a backoff mechanism includes:

[0053] The relay node state update model is constructed based on a preset relay node state update function; wherein, the expression of the relay node state update function is:

[0054]

[0055] Among them, the ζ represents the state of the i-th relay node at time t, ζ represents the control gain of the relay node, and τ represents the state of the i-th relay node at time t. t The overall adjustment target indicated by the aggregator, τ r (t) is the real-time total adjustment value;

[0056] The backoff mechanism is formulated based on a preset backoff adjustment function; wherein, the expression of the backoff adjustment function is:

[0057] rollback mode

[0058] Among them, the Represents frequency recovery output;

[0059] The convergence of the nonlinear consistency controller is controlled according to a preset convergence function; wherein, the expression of the convergence function is:

[0060] x = Θ1 = [Θ, Θ, ..., Θ] T

[0061] Wherein, 1 represents a column vector with all elements equal to 1, and Θ is the decision value that the nonlinear consistency controller aims to converge to.

[0062] The present invention constructs the relay node update model and backoff mechanism to ensure the consistency of control among each intelligent agent, and at the same time uses the convergence function to control the convergence of the nonlinear consistency controller to ensure the achievement of the control target and improve the control effect.

[0063] As a preferred example, obtaining the initial state of the large-scale isothermal controllable load cluster includes:

[0064] The initial state of the large-scale constant-temperature controllable load cluster is obtained according to preset scenario constraints; wherein, the scenario constraints include a first scenario constraint and a second scenario constraint; wherein, the expression of the first scenario constraint is:

[0065] An integral Lyapunov function is constructed based on the first scenario constraints, and the first initial state of the large-scale isothermal controllable load cluster under the first scenario constraints is obtained based on the integral Lyapunov function; wherein, the first initial state is when the state of each agent reaches the decision value of the aggregator, and the expression of the integral Lyapunov function is:

[0066]

[0067] Where s(ε) is a monotonically increasing function, and V≥0, when And only when At that time, V = 0;

[0068] when If we differentiate the integral Lyapunov function, the expression of the differentiated integral Lyapunov function is:

[0069]

[0070] in, Always true and Let be the equilibrium point of the integral Lyapunov function, where Only in x j =s(y i )-s(y j It holds true when ) = 0;

[0071] The expression for the constraint condition in the second scenario is: and Among them, when In the large-scale temperature-controlled load cluster, each temperature-controlled load no longer reduces its operating frequency, and each temperature-controlled load agent is in a second initial state; wherein, the second initial state is when the states of each temperature-controlled load agent reach consistency and Θ = 1; when At that time, each constant temperature controllable load in the large-scale constant temperature controllable load cluster no longer reduces its operating frequency, and each constant temperature controllable load intelligent agent is in the second initial state, and Θ = 0.

[0072] This invention performs different control analyses under different application scenarios, i.e., when the load is in different initial states, and tests the stability of the control based on Lyapunov stability theory, thereby further improving the control effect.

[0073] On the other hand, the present invention also discloses a constant temperature controllable load regulation system based on multi-agent control, including a thermodynamic modeling module, a control communication module, a controller construction module and a load regulation module;

[0074] The thermodynamic modeling module is used to collect the performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster, and to perform thermodynamic modeling on the room where each constant temperature controllable load is located according to the preset equivalent thermal parameter model, so as to construct the thermodynamic model corresponding to each room.

[0075] The control and communication module is used to define the large-scale constant temperature controllable load cluster as a multi-agent system, and to establish a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system.

[0076] The controller construction module is used to construct a nonlinear consistency controller based on a preset constant temperature controllable load state variable calculation method, nonlinear saturation consistency algorithm, relay node state update model and backoff mechanism.

[0077] The load control module is used to obtain the initial state of the large-scale constant temperature controllable load cluster, and to control each constant temperature controllable load according to the initial state and the thermodynamic model through the nonlinear consistency controller.

[0078] This invention discloses a constant temperature controllable load regulation system based on multi-agent control. First, the performance parameters and cooling performance indicators of each constant temperature controllable load in a large-scale constant temperature controllable load cluster are obtained. Then, a thermodynamic model corresponding to the room where the load is located is constructed according to a preset equivalent thermal parameter model. By performing thermodynamic modeling on the room where the load is located, the case of heterogeneous constant temperature controllable loads is considered, which improves the practicality of the regulation method.

[0079] Simultaneously, during the regulation process, each load is treated as an intelligent agent. A control framework for controlling the multi-agent system through an aggregator and a communication model for communication between agents are constructed, realizing unified regulation of large-scale constant-temperature controllable loads. Based on a preset constant-temperature controllable load state variable calculation method, a nonlinear saturation consensus algorithm, a relay node state update model, and a backoff mechanism, a nonlinear consensus controller is constructed. This enables unified regulation of large-scale constant-temperature controllable loads under nonlinear saturation factors when regulating the load through the nonlinear consensus controller, improving the engineering requirements capability of the method. At the same time, the backoff mechanism ensures that user experience is considered during the regulation process, improving the user experience.

[0080] As a preferred example, the thermodynamic modeling module includes a refrigeration unit, a performance unit, and a model building unit;

[0081] The refrigeration unit is used to calculate the refrigeration performance index of each constant-temperature controllable load through a preset refrigeration performance index function; wherein, the expression of the refrigeration performance index function is:

[0082]

[0083] Wherein, the γ i (t) represents the refrigeration performance index of the i-th constant-temperature controllable load, where τ i (t) represents the operating efficiency of the i-th isothermal controllable load, where ω i (t) represents the operating frequency of the i-th constant-temperature controllable load, where a i1 ,β i1 Let a be the operating power coefficient of the i-th isothermal controllable load. i2 ,β i2 Let be the coefficient of performance for the i-th constant-temperature controllable load;

[0084] The performance unit is used to calculate the performance parameters of each isothermal controllable load using a preset performance parameter function; wherein, the expression of the performance parameter function is:

[0085] H i (t)=γ i (t) / τ i (t)

[0086] Wherein, the H i (t) represents the performance parameter;

[0087] The model building unit is used to construct the thermodynamic model using a preset equivalent thermal parameter model; wherein, the expression of the thermodynamic model is:

[0088]

[0089] Where i∈π represents the set of isothermal controllable loads, t∈θ represents the set of time points, and σ i ρ represents the equivalent heat capacity of the room containing the i-th temperature-controlled load. i δ represents the equivalent thermal resistance of the room containing the i-th temperature-controlled load. i δ(t) represents the indoor temperature of the room containing the i-th constant temperature controllable load at time t, δ0(t) represents the outdoor temperature at time t, and γ i This represents the refrigeration performance index of the i-th constant temperature controllable load.

[0090] This invention calculates the performance parameters and refrigeration performance indicators to enable the subsequent construction of the thermodynamic model. This allows the actual operating conditions of the load to be considered during load regulation, thereby improving the regulation effect. At the same time, the thermodynamic model is constructed to take into account heterogeneous constant-temperature controllable loads located in different rooms, so that the regulation method provided by this invention can uniformly regulate the heterogeneous constant-temperature controllable loads, improving the practicality of regulation and adapting to different regulation scenarios. Attached Figure Description

[0091] Figure 1 This is a flowchart illustrating a method for controlling a constant temperature load based on multi-agent control, as disclosed in an embodiment of the present invention.

[0092] Figure 2 This is a schematic diagram of a constant temperature controllable load regulation system based on multi-agent control disclosed in an embodiment of the present invention.

[0093] Figure 3 This is a flowchart illustrating a method for controlling a constant temperature load based on multi-agent control, as disclosed in another embodiment of the present invention.

[0094] Figure 4This is another embodiment of the present invention, which discloses the outdoor temperature change outside the room where a large-scale constant temperature controllable load cluster is located;

[0095] Figure 5 This invention discloses a distributed circuit control framework for the standby operation of a temperature-controlled load, as another embodiment of the present invention.

[0096] Figure 6 This is a diagram showing the total power change of a temperature-controlled load multi-agent system before adopting a control strategy, as disclosed in another embodiment of the present invention.

[0097] Figure 7 This is a state change diagram of a temperature-controlled load intelligent agent under a distributed control strategy, as disclosed in another embodiment of the present invention.

[0098] Figure 8 This is a state error diagram of a temperature-controlled load intelligent agent under a distributed control strategy, as disclosed in another embodiment of the present invention.

[0099] Figure 9 This is a diagram showing the total power variation of a temperature-controlled load in an upward adjustment mode, as disclosed in another embodiment of the present invention.

[0100] Figure 10 This is a diagram showing the total power variation of a temperature-controlled load in a down-adjustment mode, as disclosed in another embodiment of the present invention. Detailed Implementation

[0101] 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.

[0102] Example 1

[0103] This embodiment discloses a method for controlling a constant temperature load based on multi-agent control. For the specific implementation process of the control method, please refer to [reference needed]. Figure 1 It mainly includes steps 101 to 104, the steps being:

[0104] Step 101: Collect the performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster, and perform thermodynamic modeling on the room where each constant temperature controllable load is located according to the preset equivalent thermal parameter model, and construct the thermodynamic model corresponding to each room.

[0105] In this embodiment, this step mainly includes: calculating the cooling performance index of each constant-temperature controllable load using a preset cooling performance index function; wherein, the expression of the cooling performance index function is:

[0106]

[0107] Wherein, the γ i (t) represents the refrigeration performance index of the i-th constant-temperature controllable load, where τ i (t) represents the operating efficiency of the i-th isothermal controllable load, where ω i (t) represents the operating frequency of the i-th constant-temperature controllable load, where a i1 ,β i1 Let a be the operating power coefficient of the i-th isothermal controllable load. i2 ,β i2 Let be the coefficient of performance for the i-th constant-temperature controllable load;

[0108] The performance parameters of each constant-temperature controllable load are calculated using a preset performance parameter function; wherein, the expression of the performance parameter function is:

[0109] H i (t)=γ i (t) / τ i (t)

[0110] Wherein, the H i (t) represents the performance parameter.

[0111] The thermodynamic model is constructed using a pre-defined equivalent thermal parameter model; wherein the expression of the thermodynamic model is:

[0112]

[0113] Where i∈π represents the set of isothermal controllable loads, t∈θ represents the set of time points, and σ i ρ represents the equivalent heat capacity of the room containing the i-th temperature-controlled load. i δ represents the equivalent thermal resistance of the room containing the i-th temperature-controlled load. i δ(t) represents the indoor temperature of the room containing the i-th constant temperature controllable load at time t, δ0(t) represents the outdoor temperature at time t, and γ i This represents the refrigeration performance index of the i-th constant temperature controllable load.

[0114] In this embodiment, this step calculates the performance parameters and refrigeration performance indicators so that the thermodynamic model can be constructed subsequently. This allows the actual operating conditions of the load to be considered when regulating the load, thereby improving the regulation effect. At the same time, the thermodynamic model is constructed to take into account heterogeneous constant temperature controllable loads located in different rooms, so that the regulation method provided by this invention can uniformly regulate the heterogeneous constant temperature controllable loads, improving the practicality of regulation and adapting to different regulation scenarios.

[0115] Step 102: Define the large-scale constant temperature controllable load cluster as a multi-agent system, and establish a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system.

[0116] In this embodiment, the step mainly includes: constructing the control framework based on the composition, communication, and room temperature of the temperature-controlled loads in the multi-agent system; wherein, the composition is that the temperature-controlled load cluster consists of heterogeneous temperature-controlled loads; the communication is that when two temperature-controlled load agents communicate in a state manner, there is a nonlinear saturation factor in the information transmission; the room temperature is that when the load adjustment time or load adjustment performance is too large, the indoor temperature of the room where the temperature-controlled load is located gradually exceeds the preset temperature comfort value; and setting several relay nodes in the multi-agent system. The relay node is defined as a communication hub; wherein the relay node communicates bidirectionally with several temperature-controlled loads in the large-scale temperature-controlled load cluster, with other adjacent temperature-controlled loads, with each relay node in pairs, and with the aggregator; the aggregator sends the preset overall load control target to the relay node connected to the aggregator, and the relay node sends the overall load control target to each agent in the multi-agent system, so that each agent can reach a consensus on the overall load control target through bidirectional communication.

[0117] In this embodiment, this step constructs the control framework based on the actual load composition, communication status, and room conditions, making the load adjustment more realistic and improving the load control effect. At the same time, a two-way communication network between intelligent agents is constructed to ensure consistent communication during load control and improve the efficiency of control.

[0118] Step 103: Construct a nonlinear consistency controller based on the preset constant temperature controllable load state variable calculation method, nonlinear saturation consistency algorithm, relay node state update model and backoff mechanism.

[0119] In this embodiment, this step mainly includes: calculating and determining the state variables of each constant-temperature controllable load based on its operating frequency information, physical constraints, application scenario requirements, and control mode; wherein, the state variables of each constant-temperature controllable load have different calculation methods for the constant-temperature controllable load state variables under the two modes of upward adjustment and downward adjustment, and the expression of the constant-temperature controllable load state variable calculation method is as follows:

[0120]

[0121] Wherein, the x i ∈x=[x1,x2,...,x n ], where n is the total number of isothermal controllable load intelligent agents, and x i (t)∈[0,1], representing the state information of the i-th agent, the It is the initial operating frequency of the i-th temperature-controlled load compressor. and These represent the lower limit and upper limit of the inherent operating frequency of the compressor with the i-th constant temperature controllable load, respectively.

[0122] The nonlinear saturated consensus algorithm is constructed based on a preset standard average consensus algorithm and a nonlinear saturation function; wherein, the expression of the standard average consensus algorithm is:

[0123]

[0124] Wherein, the Ω i It is the set of neighboring nodes of the i-th isothermal controllable load intelligent agent, wherein a ij The x is an element of the adjacency matrix A of the multi-agent communication network topology. i and x j These are the state of the i-th constant temperature controllable load agent and the state information of the j-th neighbor node of the i-th constant temperature controllable load agent, respectively.

[0125] The expression for the nonlinear saturation function is:

[0126]

[0127] in, It is a nonlinear saturation function, the y i (t) is an intermediate variable, the aforementioned n is the total number of agents;

[0128] The expression for the nonlinear saturation consensus algorithm is:

[0129]

[0130]

[0131] Wherein, ξ is the control gain of the relay node, and y i (t) is an intermediate variable;

[0132] Based on the standard average consensus algorithm and the nonlinear saturated consensus algorithm, a matrix of consensus algorithms with nonlinear protocols is obtained; wherein, the expression of the matrix is:

[0133]

[0134] in,

[0135] The relay node state update model is constructed based on a preset relay node state update function; wherein, the expression of the relay node state update function is:

[0136]

[0137] Among them, the ζ represents the state of the i-th relay node at time t, ζ represents the control gain of the relay node, and τ represents the state of the i-th relay node at time t. t The overall adjustment target indicated by the aggregator, τ r (t) is the real-time total adjustment value;

[0138] The backoff mechanism is formulated based on a preset backoff adjustment function; wherein, the expression of the backoff adjustment function is:

[0139]

[0140] Among them, the Represents frequency recovery output;

[0141] The convergence of the nonlinear consistency controller is controlled according to a preset convergence function; wherein, the expression of the convergence function is:

[0142] x = Θ1 = [Θ, Θ, ..., Θ] T

[0143] Wherein, 1 represents a column vector with all elements equal to 1, and Θ is the decision value that the nonlinear consistency controller aims to converge to.

[0144] In this embodiment, this step, based on the state variable calculation method for the isothermal controllable load, can normalize the regulation capability of heterogeneous isothermal controllable loads. Based on this, a multi-agent cooperative control method is used to control the isothermal controllable loads, thereby linking heterogeneous isothermal controllable loads using the same control strategy. Then, nonlinear saturation is combined with the standard average consensus algorithm to construct a nonlinear saturation consensus algorithm, enabling unified regulation of large-scale isothermal controllable loads under nonlinear saturation factors. This gives the method of the present invention a stronger ability to adapt to engineering needs. Furthermore, the relay node update model and backoff mechanism are constructed to ensure the consistency of control among each agent. Simultaneously, a convergence function is used to control the convergence of the nonlinear consensus controller, ensuring the achievement of the regulation target and improving the regulation effect.

[0145] Step 104: Obtain the initial state of the large-scale isothermal controllable load cluster, and regulate each isothermal controllable load according to the initial state and the thermodynamic model through the nonlinear consistency controller.

[0146] In this embodiment, the step mainly includes: obtaining the initial state of the large-scale constant-temperature controllable load cluster according to preset scenario constraints; wherein, the scenario constraints include a first scenario constraint and a second scenario constraint; wherein, the expression of the first scenario constraint is:

[0147] An integral Lyapunov function is constructed based on the first scenario constraints, and the first initial state of the large-scale isothermal controllable load cluster under the first scenario constraints is obtained based on the integral Lyapunov function; wherein, the first initial state is when the state of each agent reaches the decision value of the aggregator, and the expression of the integral Lyapunov function is:

[0148]

[0149] Where s(ε) is a monotonically increasing function, and V≥0, when And only when At that time, V = 0;

[0150] when If we differentiate the integral Lyapunov function, the expression of the differentiated integral Lyapunov function is:

[0151]

[0152] in, Always true and Let be the equilibrium point of the integral Lyapunov function, where Only in x j=s(y i )-s(y j It holds true when ) = 0;

[0153] The expression for the constraint condition in the second scenario is: and Among them, when In the large-scale temperature-controlled load cluster, each temperature-controlled load no longer reduces its operating frequency, and each temperature-controlled load agent is in a second initial state; wherein, the second initial state is when the states of each temperature-controlled load agent reach consistency and Θ = 1; when At that time, each constant temperature controllable load in the large-scale constant temperature controllable load cluster no longer reduces its operating frequency, and each constant temperature controllable load intelligent agent is in the second initial state, and Θ = 0.

[0154] In this embodiment, this step performs different control analyses under different application scenarios, i.e., when the load is in different initial states, and tests the stability of the control based on Lyapunov stability theory to further improve the control effect.

[0155] On the other hand, this embodiment also discloses a constant temperature controllable load regulation system based on multi-agent control. For the specific structural composition of the system, please refer to... Figure 2 It mainly includes a thermodynamic modeling module 201, a control and communication module 202, a controller construction module 203, and a load regulation module 204.

[0156] The thermodynamic modeling module 201 is used to collect the performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster, and to perform thermodynamic modeling on the room where each constant temperature controllable load is located according to the preset equivalent thermal parameter model, so as to construct the thermodynamic model corresponding to each room.

[0157] The control and communication module 202 is used to define the large-scale constant temperature controllable load cluster as a multi-agent system, and to establish a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system.

[0158] The controller construction module 203 is used to construct a nonlinear consistency controller based on a preset constant temperature controllable load state variable calculation method, a nonlinear saturation consistency algorithm, a relay node state update model, and a backoff mechanism.

[0159] The load control module 204 is used to obtain the initial state of the large-scale constant temperature controllable load cluster, and to control each constant temperature controllable load according to the initial state and the thermodynamic model through the nonlinear consistency controller.

[0160] In this embodiment, the thermodynamic modeling module 201 includes a refrigeration unit, a performance unit, and a model building unit.

[0161] The refrigeration unit is used to calculate the refrigeration performance index of each constant-temperature controllable load through a preset refrigeration performance index function; wherein, the expression of the refrigeration performance index function is:

[0162]

[0163] Wherein, the γ i (t) represents the refrigeration performance index of the i-th constant-temperature controllable load, where τ i (t) represents the operating efficiency of the i-th isothermal controllable load, where ω i (t) represents the operating frequency of the i-th constant-temperature controllable load, where a i1 ,β i1 Let a be the operating power coefficient of the i-th isothermal controllable load. i2 ,β i2 Let be the coefficient of performance for the i-th constant temperature controllable load.

[0164] The performance unit is used to calculate the performance parameters of each isothermal controllable load using a preset performance parameter function; wherein, the expression of the performance parameter function is:

[0165] H i (t)=γ i (t) / τ i (t)

[0166] Wherein, the H i (t) represents the performance parameter.

[0167] The model building unit is used to construct the thermodynamic model using a preset equivalent thermal parameter model; wherein, the expression of the thermodynamic model is:

[0168]

[0169] Where i∈π represents the set of isothermal controllable loads, t∈θ represents the set of time points, and σ i ρ represents the equivalent heat capacity of the room containing the i-th temperature-controlled load. i δ represents the equivalent thermal resistance of the room containing the i-th temperature-controlled load. i δ(t) represents the indoor temperature of the room containing the i-th constant temperature controllable load at time t, δ0(t) represents the outdoor temperature at time t, and γ i This represents the refrigeration performance index of the i-th constant temperature controllable load.

[0170] This embodiment discloses a method and system for regulating a constant-temperature controllable load based on multi-agent control. First, the performance parameters and cooling performance indicators of each constant-temperature controllable load in a large-scale constant-temperature controllable load cluster are obtained. Then, a thermodynamic model corresponding to the room where the load is located is constructed according to a preset equivalent thermal parameter model. By performing thermodynamic modeling on the room where the load is located, the case of heterogeneous constant-temperature controllable loads is considered, which improves the practicality of the regulation method.

[0171] Simultaneously, during the regulation process, each load is treated as an intelligent agent. A control framework for controlling the multi-agent system through an aggregator and a communication model for communication between agents are constructed, realizing unified regulation of large-scale constant-temperature controllable loads. Based on a preset constant-temperature controllable load state variable calculation method, a nonlinear saturation consensus algorithm, a relay node state update model, and a backoff mechanism, a nonlinear consensus controller is constructed. This enables unified regulation of large-scale constant-temperature controllable loads under nonlinear saturation factors when regulating the load through the nonlinear consensus controller, improving the engineering requirements capability of the method. At the same time, the backoff mechanism ensures that user experience is considered during the regulation process, improving the user experience.

[0172] Example 2

[0173] This embodiment also discloses a method for controlling a constant temperature load based on multi-agent control. The specific implementation process of the control method can be referred to... Figure 3 It mainly includes steps 301 to 303, the steps being:

[0174] Step 301: Perform operational modeling for each constant temperature controllable load in the large-scale constant temperature controllable load cluster to be regulated, and perform thermodynamic modeling for the corresponding room based on the first-order equivalent thermal parameters, and set the performance parameters and cooling performance indicators for each constant temperature controllable load; wherein, the performance parameters are related to the working power and cooling performance indicators of the constant temperature controllable load.

[0175] In this embodiment, the main steps are: collecting the performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster, and performing thermodynamic modeling on the room where each constant temperature controllable load is located according to the preset equivalent thermal parameter model, and constructing the thermodynamic model corresponding to each room.

[0176] Specifically, in this embodiment, a room thermodynamic model is established, wherein the expression of the room thermodynamic model is:

[0177]

[0178] Where i∈π represents the set of isothermal controllable loads, t∈θ represents the set of time points, and σ iρ represents the equivalent heat capacity of the room containing the i-th temperature-controlled load. i δ represents the equivalent thermal resistance of the room containing the i-th temperature-controlled load. i δ(t) represents the indoor temperature of the room containing the i-th constant temperature controllable load at time t, δ0(t) represents the outdoor temperature at time t, and γ i This represents the refrigeration performance index of the i-th constant temperature controllable load.

[0179] The refrigeration performance index is defined as follows:

[0180]

[0181] Wherein, the γ i (t) represents the refrigeration performance index of the i-th constant-temperature controllable load, where τ i (t) represents the operating efficiency of the i-th isothermal controllable load, where ω i (t) represents the operating frequency of the i-th constant-temperature controllable load, where a i1 ,β i1 Let a be the operating power coefficient of the i-th isothermal controllable load. i2 ,β i2 Let be the coefficient of performance for the i-th constant temperature controllable load.

[0182] Define performance parameters, the expressions for which are:

[0183] H i (t)=γ i (t) / τ i (t)

[0184] Wherein, the H i (t) represents the performance parameter.

[0185] Preferably, in this embodiment, referring to Figure 5 Taking the community where the distributed circuit control framework for the operation backup of the temperature-controlled load is shown as an example, kinematic and dynamic modeling is performed on all temperature-controlled loads, and a parameter table of the temperature-controlled load and its room is set. In this embodiment, the following parameter table is considered:

[0186] Table 1. Temperature-Controllable Loads and Their Room Parameters

[0187]

[0188] Where α i1 ,β i1 α i2 ,β i2 , σ i , ρi δ i These represent the operating power coefficient, cooling performance coefficient, lower operating frequency limit, upper operating frequency limit, equivalent heat capacity, equivalent thermal resistance, and indoor temperature setpoint of the i-th constant temperature controllable load, respectively. Additionally, the outdoor temperature variation of the room containing the constant temperature controllable load over 24 hours is set as follows: Figure 4 As shown.

[0189] Next, a thermodynamic model is created for the room containing the constant temperature controllable load based on the performance indicators, performance parameters, and the thermodynamic modeling process.

[0190] Step 302: Treat the large-scale constant temperature controllable load cluster as a multi-agent system, and establish a control framework and a communication model between agents to control the multi-agent system through an aggregator.

[0191] In this embodiment, the main steps are: defining the large-scale constant temperature controllable load cluster as a multi-agent system, and establishing a control framework for controlling the multi-agent system through an aggregator and a communication model between each pair of agents in the multi-agent system.

[0192] Specifically, in this embodiment, the control framework considers different situations corresponding to the large-scale temperature-controlled load cluster, which mainly includes load composition, communication, and regulation performance. The load composition is as follows: the temperature-controlled load cluster is composed of heterogeneous temperature-controlled loads. The temperature-controlled load cluster is regarded as a multi-agent system, and each temperature-controlled load is one of the agents.

[0193] The communication situation is as follows: the communication between constant temperature controllable loads is the communication between agents in a multi-agent system. When two constant temperature controllable load agents communicate in a state, there is a nonlinear saturation factor in the information transmission.

[0194] Regarding the adjustment performance: Due to the thermal inertia of the building, the indoor temperature changes slowly. When the adjustment time or adjustment performance is too large, the indoor temperature will gradually exceed the user's predetermined comfort value.

[0195] Referring to the various scenarios included in the aforementioned control framework, several relay nodes are set up in the multi-agent system composed of temperature-controlled loads. The relay nodes themselves do not perform load control but act as communication hubs, facilitating information exchange between the aggregator and the temperature-controlled load agent cluster. Two-way communication is possible between relay nodes and several temperature-controlled loads, between temperature-controlled loads and other adjacent temperature-controlled loads, between each relay node, and between relay nodes and the aggregator to exchange information. When control is required, the aggregator provides the corresponding overall load control target and sends the control target to the relay node connected to it. The various agents then reach a consensus on the control target through bidirectional communication on the communication network, thereby realizing multi-agent collaborative control. The aggregator is a server or other equipment used to aggregate a certain number of temperature-controlled loads for large-scale load regulation. In the control framework, the aggregator undertakes the roles of control and communication. Specifically, the aggregator communicates with each relay node, and the relay nodes communicate with the temperature-controlled loads. When load regulation is required, the power grid dispatcher sends dispatch information to the aggregator, which sends the regulation command to the relay node. The load regulation is then completed through communication between the relay node and the temperature-controlled load.

[0196] Step 303: Design a nonlinear uniform controller and, in conjunction with the constructed thermodynamic model, regulate the large-scale isothermal controllable load cluster.

[0197] In this embodiment, the main steps are as follows: constructing a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consistency algorithm, a relay node state update model, and a fallback mechanism; obtaining the initial state of the large-scale constant-temperature controllable load cluster; and regulating each constant-temperature controllable load through the nonlinear consistency controller based on the initial state and the thermodynamic model.

[0198] Specifically, in this embodiment, during the consistency control process, the information exchanged between adjacent agents contains individual state variable information. Based on the operating frequency information, physical constraints, application scenario requirements, and control mode of the temperature-controlled load, the state variables of each agent can be derived. The state variables of each temperature-controlled load agent have different calculation methods under both upward and downward adjustment modes, as shown in the following formula:

[0199]

[0200] Wherein, the x i ∈x=[x1,x2,...,x n ], where n is the total number of isothermal controllable load intelligent agents, and x i(t)∈[0,1], representing the state information of the i-th agent, the It is the initial operating frequency of the i-th temperature-controlled load compressor. and These represent the lower limit and upper limit of the inherent operating frequency of the compressor with the i-th constant temperature controllable load, respectively.

[0201] The calculation formula for the state variables can be used to normalize the controllability of heterogeneous isothermal controllable loads, and based on this, a multi-agent cooperative control method can be used to control the isothermal controllable loads. This is the theoretical basis for linking heterogeneous isothermal controllable loads using the same control strategy.

[0202] Next, the nonlinear saturated consensus algorithm is constructed using the standard average consensus algorithm based on undirected graphs and a nonlinear saturation function. The expression of the standard average consensus algorithm based on undirected graphs is as follows:

[0203]

[0204] Wherein, the Ω i It is the set of neighboring nodes of the i-th isothermal controllable load intelligent agent, wherein a ij The x is an element of the adjacency matrix A of the multi-agent communication network topology. i and x j These are the state of the i-th constant temperature controllable load agent and the state information of the j-th neighbor node of the i-th constant temperature controllable load agent, respectively.

[0205] Furthermore, the information exchange between two adjacent agents exhibits a nonlinear saturation factor. In the model, the nonlinear saturation function in the information exchange process between isothermal controllable loads is specifically represented by the following formula:

[0206]

[0207] in, It is a nonlinear saturation function, the y i (t) is an intermediate variable, the aforementioned The number n is the total number of agents.

[0208] By incorporating the nonlinear saturated standard average consensus algorithm, a nonlinear saturated consensus algorithm is constructed, as shown in the following equation:

[0209]

[0210]

[0211] In the above formula, ξ is the control gain of the relay node, and y i(t) is an intermediate variable.

[0212] Furthermore, by combining the standard average consensus algorithm and the nonlinear saturation consensus algorithm, a matrix of consensus algorithms with nonlinear protocols is obtained, and the expression of the rectangle is as follows:

[0213]

[0214] in,

[0215] Furthermore, the state update expression for the relay node is as follows:

[0216]

[0217] Among them, the ζ represents the state of the i-th relay node at time t, ζ represents the control gain of the relay node, and τ represents the state of the i-th relay node at time t. t The overall adjustment target indicated by the aggregator, τ r (t) is the real-time total adjustment value.

[0218] Furthermore, if the adjustment time or adjustment performance is too large, causing the indoor temperature to gradually exceed the user's preset comfort value, a fallback mechanism will be triggered. Taking the above adjustment mode as an example, it is shown in the following formula:

[0219]

[0220] in, Represents frequency recovery output; This represents the lower limit of the inherent operating frequency of the compressor for the i-th constant temperature controllable load; when the consistency control output is disabled, the frequency recovery output will be used.

[0221] Furthermore, in the control of isothermal controllable loads based on multi-agent cooperative control, the convergence of the consensus algorithm is a crucial issue because it directly relates to the achievement of the control objective. Assuming the control algorithm aims to achieve the following objectives, the control convergence algorithm is as follows:

[0222] x = Θ1 = [Θ, Θ, ..., Θ] T

[0223] Where 1 represents a column vector with all elements equal to 1; Θ is the decision value that the algorithm in this paper aims to converge to.

[0224] In this step, the implementation of constant temperature controllable load regulation based on multi-agent cooperative control can be analyzed in two scenarios according to different initial states:

[0225] Scenario 1: Satisfying season in Based on scenario one, construct the following integral Lyapunov function as follows:

[0226]

[0227] in It is a monotonically increasing function, and V≥0; furthermore, for an undirected graph... In this context, its Laplace matrix L is a symmetric matrix with eigenvalues ​​of 0 and corresponding eigenvectors of 1. Therefore, since Based on the above formula, if and only if When V = 0, then the fundamental requirement for the Lyapunov function is satisfied. Because... Therefore, the derivative of the Lyapunov function is as follows:

[0228]

[0229] Obviously, It always holds true, and It is the Lyapunov equilibrium point, in addition Only This holds true at that time. In other words, under a given scenario, the agent's state can reach the aggregator's decision value, i.e.

[0230] Scenario 2: Satisfying and When; when All temperature-controlled loads no longer reduce their operating frequency, therefore the state consistency of the agent can be achieved, and x = 1, i.e., Θ = 1; when At this time, all temperature-controlled loads no longer increase their operating frequency, so the state consistency of the agent can be achieved, and x = 0, that is, Θ = 0.

[0231] Preferably, in this embodiment, referring to the parameters of the temperature-controlled load and its room provided in step 301, the control parameters ζ = 0.002 and ξ = 0.15 are given. Then, the power variation of the temperature-controlled load multi-agent system over 24 hours without a control strategy can be referenced. Figure 6 When the nonlinear consistency controller provided in this embodiment is used for regulation, the state change of the isothermal controllable load agent under the distributed control strategy can be referred to Figure 7 During regulation, the state error between the state of the isothermal controllable load agent and the final decision value can be referenced. Figure 8 When adjusting the load, the total power change of the constant-temperature controllable load in the upward and downward adjustment modes can be referred to respectively. Figure 9 and Figure 10 ,thus, Figure 7and Figure 8 This shows the state change under the consideration of nonlinear saturation, thus proving that the consistency of the constant temperature controllable load state is perfectly achieved under the nonlinear saturation factor. Figure 9 and Figure 10 This also proves the effectiveness of the two modes of state regulation mechanism designed in this patent: upward and downward regulation.

[0232] This embodiment provides a method for regulating a temperature-controlled load based on multi-agent control, which is more practical and considers the situation of heterogeneous temperature-controlled loads. It proposes a standardized method for quantifying heterogeneous temperature-controlled loads, making the designed control method more practical and engineering-significant. A backoff mechanism is proposed, which is activated when the temperature is within the user's expected range, thus significantly improving the user experience. Furthermore, it achieves unified regulation of large-scale temperature-controlled loads under nonlinear saturation factors, giving the method greater adaptability to engineering needs. In addition, the consistency control algorithm of this invention has been proven based on Lyapunov stability theory.

[0233] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for controlling a temperature-controlled load based on multi-agent control, characterized in that, include: The performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster are collected, and the room where each constant temperature controllable load is located is thermodynamically modeled according to the preset equivalent thermal parameter model, and the thermodynamic model corresponding to each room is constructed. The large-scale constant temperature controllable load cluster is defined as a multi-agent system, and a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system are established. A nonlinear consensus controller is constructed based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consensus algorithm, a relay node state update model, and a backoff mechanism. Specifically, the state variables of each constant-temperature controllable load are determined based on its operating frequency information, physical constraints, application scenario requirements, and control mode. A nonlinear saturated consensus algorithm is constructed based on a preset standard average consensus algorithm and a nonlinear saturation function. A matrix of consensus algorithms with nonlinear protocols is obtained based on the standard average consensus algorithm and the nonlinear saturated consensus algorithm. A relay node state update model is constructed based on a preset relay node state update function. A backoff mechanism is formulated based on a preset backoff adjustment function. A convergence function is set to control the convergence of the nonlinear consistency controller; wherein, several relay nodes are set in the multi-agent system, and the relay nodes are defined as communication hubs; wherein, the relay nodes communicate bidirectionally with several temperature-controlled loads in the large-scale temperature-controlled load cluster, with the temperature-controlled loads and other adjacent temperature-controlled loads, with each pair of relay nodes, and with the aggregator; when the load adjustment time or load adjustment performance is too large, causing the indoor temperature of the room where the temperature-controlled load is located to gradually exceed the preset temperature comfort value, a fallback mechanism is triggered to control the temperature-controlled load to switch to fallback mode; in the fallback mode, the operating frequency of the temperature-controlled load is set to frequency recovery output; The initial state of the large-scale isothermal controllable load cluster is obtained, and each of the isothermal controllable loads is regulated by the nonlinear consistency controller based on the initial state and the thermodynamic model.

2. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The collection of performance parameters and cooling performance indicators for each constant-temperature controllable load in the large-scale constant-temperature controllable load cluster includes: The cooling performance index of each constant-temperature controllable load is calculated using a preset cooling performance index function; wherein the expression of the cooling performance index function is: Among them, the For the first The refrigeration performance indicators of a constant temperature controllable load, the For the first The working efficiency of a constant temperature controllable load, the For the first The operating frequency of a constant temperature controllable load, the For the first The operating power coefficient of a constant temperature controllable load, the For the first The coefficient of performance (COP) of a constant-temperature controllable load; wherein, the The set representing a temperature-controlled load, the A set representing points in time; The performance parameters of each constant-temperature controllable load are calculated using a preset performance parameter function; wherein, the expression of the performance parameter function is: Among them, the The performance parameters are as described above.

3. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The process involves performing thermodynamic modeling on each room containing a constant-temperature controllable load based on a preset equivalent thermal parameter model, constructing a thermodynamic model corresponding to each room, including: The thermodynamic model is constructed using a pre-defined equivalent thermal parameter model; wherein the expression of the thermodynamic model is: Among them, the This represents the set of temperature-controlled loads, the... The set representing time points, the Indicates the first The equivalent heat capacity of a room containing a constant-temperature controllable load, the Indicates the first The equivalent thermal resistance of the room containing a constant-temperature controllable load, the Indicates the first The room containing the constant temperature and controllable load is at a certain time. The indoor temperature, the Indicates time The outdoor temperature, the Indicates the first The refrigeration performance indicators of a constant temperature controllable load; wherein, the The set representing a temperature-controlled load, the A set representing points in time.

4. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The establishment of a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system includes: The control framework is constructed based on the composition, communication, and room temperature of the temperature-controlled loads in the multi-agent system. The composition involves a temperature-controlled load cluster composed of heterogeneous temperature-controlled loads. The communication involves nonlinear saturation factors in information transmission when two temperature-controlled load agents communicate. The room temperature involves the gradual increase in indoor temperature beyond the preset comfort value when the load adjustment time or performance is too high. The aggregator sends the preset overall load control target to the relay node connected to the aggregator, and the relay node sends the overall load control target to each agent in the multi-agent system, so that each agent can reach a consensus on the overall load control target through bidirectional communication.

5. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The construction of a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consistency algorithm, a relay node state update model, and a backoff mechanism includes: The state variables of each temperature-controlled load are calculated and determined based on its operating frequency information, physical constraints, application scenario requirements, and control mode. The state variables of each temperature-controlled load have different calculation methods under both upward and downward adjustment modes. The expression for the calculation method of the temperature-controlled load state variables is as follows: Among them, the The It is the total number of temperature-controlled load intelligent agents, the aforementioned (t) ∈ [0,1] represents the th The state information of each agent, the It is the first The initial operating frequency of a compressor with a constant temperature and controllable load, the and Representing the first The inherent lower limit and inherent upper limit of the operating frequency of a compressor with constant temperature and controllable load; For the first The operating frequency of a constant temperature controllable load at time t.

6. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The construction of a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consistency algorithm, a relay node state update model, and a backoff mechanism includes: The nonlinear saturated consensus algorithm is constructed based on a preset standard average consensus algorithm and a nonlinear saturation function; wherein, the expression of the standard average consensus algorithm is: Among them, the It is the first The set of neighboring nodes of a temperature-controlled load intelligent agent, the It is the adjacency matrix of the multi-agent communication network topology. The element, the and They are the first The state of the temperature-controlled load intelligent agent and the first The first temperature-controlled load intelligent agent The status information of each neighboring node; The set representing a temperature-controlled load, the A set representing points in time; The expression for the nonlinear saturation function is: in, , is a nonlinear saturation function, the It is an intermediate variable. Where n is the total number of agents; The expression for the nonlinear saturation consensus algorithm is: Among them, the It is the control gain of the relay node, the It is an intermediate variable; the stated This refers to the state information of multiple agents when they reach consensus at time t. Based on the standard average consensus algorithm and the nonlinear saturated consensus algorithm, a matrix of consensus algorithms with nonlinear protocols is obtained; wherein, the expression of the matrix is: in, Wherein, L is a feature vector with a feature value of 0 and the feature value corresponding to that feature is... A symmetric matrix.

7. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The construction of a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consistency algorithm, a relay node state update model, and a backoff mechanism includes: The relay node state update model is constructed based on a preset relay node state update function; wherein, the expression of the relay node state update function is: Among them, the Representing the Each relay node The state at that moment, the state described The control gain representing the relay node, the The overall adjustment target indicated by the aggregator, the It is a real-time adjustment of the total value; the aforementioned The set representing a temperature-controlled load, the The set representing points in time; It is the control gain of the relay node; the It is the adjacency matrix of the multi-agent communication network topology. Element; The backoff mechanism is formulated based on a preset backoff adjustment function; wherein, the expression of the backoff adjustment function is: Among them, the Represents frequency recovery output; The convergence of the nonlinear consistency controller is controlled according to a preset convergence function; wherein, the expression of the convergence function is: Among them, the Represents a column vector where all elements are 1. It is the decision value that the nonlinear consensus controller aims to converge to; wherein, the represent In the moment The state of a constant-temperature controllable load intelligent agent; the represent In the moment The first temperature-controlled load intelligent agent The status information of each neighboring node; For the first The operating frequency of a constant temperature controllable load; For the first The operating frequency of a constant-temperature controllable load at time t; Representing the The initial operating frequency of a compressor with a constant temperature and controllable load; where x represents an intelligent agent.

8. The method for controlling a constant temperature load based on multi-agent control as described in claim 1, characterized in that, The process of obtaining the initial state of the large-scale constant-temperature controllable load cluster includes: The initial state of the large-scale constant-temperature controllable load cluster is obtained according to preset scenario constraints; wherein, the scenario constraints include a first scenario constraint and a second scenario constraint; wherein, the expression of the first scenario constraint is: ; ; ; An integral Lyapunov function is constructed based on the first scenario constraints, and the first initial state of the large-scale isothermal controllable load cluster under the first scenario constraints is obtained based on the integral Lyapunov function; wherein, the first initial state is when the state of each agent reaches the decision value of the aggregator, and the expression of the integral Lyapunov function is: in, It is a monotonically increasing function, and V 0, when =-L( 1)=- L1=0 and only if y= When 1=x, V=0, the aforementioned Intermediate variables representing all agents The average of the initial values; If we differentiate the integral Lyapunov function, the expression of the differentiated integral Lyapunov function is: in, Always true and Let be the equilibrium point of the integral Lyapunov function, where Only When established; the aforementioned The set representing a temperature-controlled load, the The set representing points in time; It is the adjacency matrix of the multi-agent communication network topology. Element; The expression for the constraint condition in the second scenario is: Among them, when In the large-scale temperature-controlled load cluster, each temperature-controlled load no longer reduces its operating frequency, and each temperature-controlled load agent is in a second initial state; wherein, the second initial state is when the states of each temperature-controlled load agent reach a consensus and ;when At that time, each temperature-controlled load in the large-scale temperature-controlled load cluster no longer reduces its operating frequency, and each temperature-controlled load agent is in the second initial state. ; wherein, the The nonlinear consensus controller is to converge to the decision value; n represents the total number of agents; V represents the sign of the integral Lyapunov function; L is a variable with a zero eigenvalue and the corresponding eigenvector is... A symmetric matrix; where x represents the agent; the... The intermediate variable representing the initial moment; the and They are the first The state of the temperature-controlled load intelligent agent and the first The first temperature-controlled load intelligent agent The state information of each neighboring node.

9. A constant-temperature controllable load regulation system based on multi-agent control, characterized in that, It includes a thermodynamic modeling module, a control and communication module, a controller construction module, and a load regulation module; The thermodynamic modeling module is used to collect the performance parameters and cooling performance indicators of each constant temperature controllable load in the large-scale constant temperature controllable load cluster, and to perform thermodynamic modeling on the room where each constant temperature controllable load is located according to the preset equivalent thermal parameter model, so as to construct the thermodynamic model corresponding to each room. The control and communication module is used to define the large-scale constant temperature controllable load cluster as a multi-agent system, and to establish a control framework for controlling the multi-agent system through an aggregator and a communication model between pairs of agents in the multi-agent system. The controller construction module is used to construct a nonlinear consistency controller based on a preset method for calculating the state variables of a constant-temperature controllable load, a nonlinear saturated consensus algorithm, a relay node state update model, and a backoff mechanism. Specifically, it calculates and determines the state variables of each constant-temperature controllable load based on its operating frequency information, physical constraints, application scenario requirements, and control mode; constructs the nonlinear saturated consensus algorithm based on a preset standard average consensus algorithm and a nonlinear saturation function; obtains a matrix of consensus algorithms with a nonlinear protocol based on the standard average consensus algorithm and the nonlinear saturated consensus algorithm; constructs the relay node state update model based on a preset relay node state update function; and formulates a backoff adjustment mechanism based on a preset backoff adjustment function. The system includes a fallback mechanism; it controls the convergence of the nonlinear consistency controller according to a preset convergence function; wherein, several relay nodes are set in the multi-agent system, and the relay nodes are defined as communication hubs; wherein, the relay nodes communicate bidirectionally with several temperature-controlled loads in the large-scale temperature-controlled load cluster, with the temperature-controlled loads and other adjacent temperature-controlled loads, with each pair of relay nodes, and with the aggregator; when the load adjustment time or load adjustment performance is too large, causing the indoor temperature of the room where the temperature-controlled load is located to gradually exceed the preset temperature comfort value, the fallback mechanism is triggered, controlling the temperature-controlled load to switch to fallback mode; in the fallback mode, the operating frequency of the temperature-controlled load is set to frequency recovery output; The load control module is used to obtain the initial state of the large-scale constant temperature controllable load cluster, and to control each constant temperature controllable load according to the initial state and the thermodynamic model through the nonlinear consistency controller.

10. The constant temperature controllable load regulation system based on multi-agent control as described in claim 9, characterized in that, The thermodynamic modeling module includes a refrigeration unit, a performance unit, and a model building unit; The refrigeration unit is used to calculate the refrigeration performance index of each constant-temperature controllable load through a preset refrigeration performance index function; wherein, the expression of the refrigeration performance index function is: Among them, the For the first The refrigeration performance indicators of a constant temperature controllable load, the For the first The working efficiency of a constant temperature controllable load, the For the first The operating frequency of a constant temperature controllable load, the For the first The operating power coefficient of a constant temperature controllable load, the For the first The coefficient of performance (COP) of a constant-temperature controllable load; The set representing a temperature-controlled load, the A set representing points in time; The performance unit is used to calculate the performance parameters of each isothermal controllable load using a preset performance parameter function; wherein, the expression of the performance parameter function is: Among them, the The performance parameters are as described above; The model building unit is used to construct the thermodynamic model using a preset equivalent thermal parameter model; wherein, the expression of the thermodynamic model is: Among them, the This represents the set of temperature-controlled loads, the... The set representing time points, the Indicates the first The equivalent heat capacity of a room containing a constant-temperature controllable load, the Indicates the first The equivalent thermal resistance of the room containing a constant-temperature controllable load, the Indicates the first The room containing the constant temperature and controllable load is at a certain time. The indoor temperature, the Indicates time The outdoor temperature, the Indicates the first The refrigeration performance indicators of a constant temperature controllable load; The set representing a temperature-controlled load, the A set representing points in time.

Images