Central air conditioning system self-adaptive control method and system considering time lag effect and storage medium

Abstract

The invention discloses a central air-conditioning system self-adaptive control method and system considering the time-lag effect and a storage medium, and the method comprises the steps that a central air-conditioning system time-lag state space model based on energy balance is constructed to describe the time-lag phenomenon of a central air-conditioning system; through data fitting and simulated annealing algorithm optimization, identifying a time lag value of the central air-conditioning system; based on the time-lag value, the time-lag state space model of the central air-conditioning system is applied to predict a time-lag stable interval of the central air-conditioning system; and according to the time-delay stable interval, the central air conditioning system is controlled to operate in a self-adaptive mode by combining fuzzy PID. According to the method, the time lag phenomenon in the system can be accurately described, the control precision and energy efficiency of the system can be improved, the problem of control lag of the central air conditioning system in the demand response process can be solved, the overall operation efficiency of the system is improved, and the method has important theoretical significance and practical application value.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioning control, and in particular to a central air conditioning system adaptive control method and system, and a storage medium taking into account time lag effects. Background Art

[0002] As an indispensable infrastructure in modern large buildings, central air conditioning systems are widely used in office buildings, shopping malls, hospitals, and residences. Their main function is to adjust indoor temperature and humidity by cooling and heating the air, thereby meeting people's needs for a comfortable environment. The operating efficiency of the central air conditioning system is directly related to the energy consumption and operating costs of the building. Therefore, it is very meaningful to improve the control accuracy and energy-saving effect of the central air conditioning system.

[0003] Traditional central air-conditioning system control methods mainly rely on experience and simple linear models. Common control strategies such as PID control, fuzzy control, expert system control, etc., can achieve stable operation of the system to a certain extent. However, the time lag effect that is prevalent in central air-conditioning systems is often ignored or simply handled, resulting in delayed system control response and reduced adjustment accuracy, which in turn affects the overall energy efficiency of the system. The time lag effect refers to the phenomenon that after the system receives the input signal, the output signal has a certain time delay. This delay may be caused by various factors such as air transmission, equipment inertia, sensor delay or data processing delay. In large buildings, central air-conditioning systems usually have complex thermodynamic characteristics and multivariable coupling relationships, which makes the time lag effect more significant. For example, when the air-conditioning system adjusts the supply air temperature or flow, the change in indoor temperature does not respond immediately, but gradually manifests itself after a period of time. If this time lag phenomenon is not handled properly, the control system will be unable to adjust the operating parameters in time, resulting in overcooling or heating, resulting in unnecessary energy waste. In addition, central air-conditioning systems are usually composed of multiple subsystems, such as chillers, air handling units, duct systems, etc. There are complex interactions between each subsystem, which further aggravates the system's time lag problem.

[0004] With the rapid development of building information and intelligent technology, more and more data are collected and applied to the control and optimization of central air-conditioning systems. Building energy management systems (BEMS) can store and process a large amount of historical operation data, which provides researchers with an opportunity to deeply analyze the time-lag effect of central air-conditioning systems. However, due to the long data collection cycle, the continuity and integrity of system data are subject to certain limitations, making the identification and modeling of time-lag phenomena more complex and difficult.

[0005] In response to the above problems, domestic and foreign scholars have proposed a variety of improvement methods in recent years, such as applying intelligent algorithms, neural networks and adaptive control technologies to enhance the control ability of the system, especially in dealing with time-delay problems. For example, the particle swarm optimization algorithm and Kalman filter technology are used to identify the time-delay value in the system, and the accuracy of the control model is improved through data-driven methods; predictive control and foresight control technologies are used to feed back future information to the current control operation to alleviate the adverse effects of the time-delay effect. However, these methods still have problems such as high model complexity, large amount of calculation, and poor real-time performance in specific applications, and further verification and optimization are needed in actual engineering.

[0006] In view of the above problems and limitations, it is an urgent problem to propose a central air-conditioning demand response modeling method that can fully consider the time lag effect. Summary of the invention

[0007] In view of this, the present invention provides a central air-conditioning system adaptive control method and system, and a storage medium taking into account the time lag effect, so as to improve the control accuracy and energy efficiency of the central air-conditioning system.

[0008] The technical solution of the present invention to solve the above technical problem is as follows: a central air conditioning system adaptive control method considering time lag effect, comprising:

[0009] Construct a time-delay state-space model of central air-conditioning system based on energy balance to describe the time-delay phenomenon of central air-conditioning system;

[0010] Through data fitting and simulated annealing algorithm optimization, the time lag value of the central air conditioning system is identified;

[0011] Based on the time lag value, applying the central air conditioning system time lag state space model to predict the time lag stability interval of the central air conditioning system;

[0012] Based on the time-delay stable interval, the operation of the central air-conditioning system is controlled by combining fuzzy PID adaptive control;

[0013] The time-delay state space model of the central air conditioning system is expressed as follows:

[0014]

[0015] Where x(t) is the n-dimensional state vector x(t) = [x1(t), x2(t), ..., x n (t)] T , including the state parameters of temperature, pressure and flow rate; u(t) is the n-dimensional input vector u(t) = [u1(t),u2(t),...,u n (t)] T, including inlet temperature, inlet pressure, inlet flow, and control opening; y(t) is the m-dimensional output vector y(t) = [y1(t), y2(t), ..., y n (t)] T , including the state parameters of outlet temperature, outlet pressure, outlet flow, and opening output; A and B are the coefficient matrix and input matrix respectively, C and D are the output matrix and direct transmission matrix respectively;

[0016] Preferably, the construction of a time-delay state space model of a central air-conditioning system based on energy balance includes:

[0017] 1) Determine the basic equilibrium equation:

[0018] The change in thermodynamic state is the sum of the energy absorbed or released by the medium and the energy exchanged with the outside world:

[0019]

[0020] Where T is temperature, °C; t is time, s; c is mass specific heat capacity, J / (kg·°C); a is heat transfer coefficient, W / (m 2 ·℃); ΔT1 and ΔT2 are the temperature differences after heat absorption and heat exchange, respectively, ℃;

[0021] 2) Linearize the basic equilibrium equation and convert it into incremental form: Re-express the basic variables in the equation as the sum of the steady-state initial value and a small increment; linearize the coefficients in the equation using the first-order Taylor expansion;

[0022] 3) Determine variables: specify the inlet parameters of the central air-conditioning system as input variables, and the outlet parameters of the central air-conditioning system as output variables; determine the state variables according to the modeling requirements;

[0023] 4) Forming the coefficient matrix: Analyze the increment and variable type and extract relevant quantities from the increment representation;

[0024] Convert the cumulative form into a matrix product, and finally integrate multiple equilibrium equations into a unified matrix form;

[0025] 5) Convert to standard form and use the coefficient matrix to write the normalized state space equations and output equations.

[0026] Preferably, the time lag problem is converted into an optimization problem under constraints by using the data fitting method; and the minimum time lag value under the constraints is searched for by using a simulated annealing algorithm.

[0027] Preferably, the method of converting the time lag problem into an optimization problem under constraints by means of data fitting includes:

[0028] Obtain two correlated state parameter change curves in the central air-conditioning system and compare them; use the minimum-maximum normalization method to map the data to the [0,1] interval, and the formula is as follows:

[0029]

[0030] In the formula, max is the maximum value of the sample data, min is the minimum value of the sample data, and P is the conversion data; g1 (t-τ) and P g2 The mean of the absolute value of the derivative difference of (t) integrated over the range (t1, t2) is defined as P g1 With P g2 The trend of the time lag between τ and τ is the actual time lag of the system; the expression is:

[0031]

[0032] The size of the time lag is comprehensively considered through the changing trend within the same daily sampling period; P1,P2 The smaller it is, the more consistent the trends of P1 and P2 are, and the smaller the hysteresis value is. When the P2 state parameter changes for the first time, the hysteresis is considered to end, and the actual hysteresis is t P1,P2 The value corresponding to the minimum value is

[0033] τ=τ min =argmint P1,P2 (5)

[0034] The corresponding time-lag series is

[0035] τ=[τ1,τ2,...,τ N ] (6)

[0036] Among them, as the state reference time point, τ N =0, indicating that the time lag of the air outlet variable at the last moment is 0, that is, the zero time lag reference variable; τ1~τ N-1 In turn, they represent the time lag between the zero-time-lag reference variable and other corresponding groups of key state variables; constructing the trend similarity matrix between multiple groups of data, the expression is:

[0037]

[0038] In the formula, i = 1, 2, ..., N-1; tp reflects the trend consistency among multiple variables;

[0039] The L2 norm is used to quantify the trend vector, and the minimum value of the L2 norm of tp||tp||2 is obtained as the time lag value of the link described by the established state association data matrix. The expression is:

[0040] γ i=min(‖tp‖2) (8)

[0041] The τ obtained at this time may be a suboptimal solution. Further optimization calculations are performed to convert the time-delay problem into an optimization problem under constraints.

[0042]

[0043] Preferably, the method of searching for the minimum hysteresis value under the constraint conditions by using the simulated annealing algorithm includes: the control parameter is temperature T, the objective function is internal energy E, and the state node at temperature T is x i ; When the control parameter T decreases, the internal energy E of the objective function will also decrease until the temperature reaches the lowest energy state. At this time, the global minimum value searched is the optimal value of the data trend similarity matrix.

[0044] Preferably, the method of predicting the time-delay stability interval of the central air-conditioning system by applying the time-delay state space model of the central air-conditioning system based on the time-delay value includes:

[0045] The integral inequality method is used to establish the stability criterion related to the time-delay range of the time-delay state space model prediction of the central air-conditioning system. Let the change speed of the time-delay value be u, the lower bound of the time-delay stability be τ1, and the upper bound be τ2.

[0046] e i is the block coordinate matrix (i=1,...,9),

[0047] For a given scalar 0<τ1<τ2, 0≤μ<1, if there exists a matrix Q of suitable dimension i >0(i=1,...,5), Z j >0(j=1,...,4), R1>0, R2>0, P=[P ij ] 5×5 >0, the following linear matrix inequality holds, and the model is asymptotically stable:

[0048]

[0049] in,

[0050]

[0051] Preferably, the operation of the central air-conditioning system is controlled based on the time-delay stable interval in combination with fuzzy PID adaptive control, including:

[0052] Fuzzy PID adaptive control converts the fuzzy regions of related parameters such as hot and cold temperatures, flow rate, and running time into fuzzy subsets between 0 and 1 through membership functions to describe the current state of the system.

[0053] Fuzzy reasoning controls the output based on the real-time status information of the central air-conditioning system and the fuzzy reasoning rules in the database, and processes the fuzzy control value precisely, directly acting on the actuator of the air-conditioning system, thereby realizing dynamic adjustment of the system.

[0054] Preferably, a Fuzzy-PID controller is used in the fuzzy PID adaptive control, wherein the algorithm expression of the PID controller is:

[0055]

[0056] In the formula, k p is the proportionality coefficient; T i is the integration time constant; T d is the differential time constant; e(t) is the control deviation e(t) = r(t)-y(t).

[0057] In a second aspect, the present invention further provides an adaptive control system for a central air conditioning system taking into account a time lag effect, comprising:

[0058] A module for building a time-delay state-space model of a central air-conditioning system based on energy balance, which is used to describe the time-delay phenomenon of the central air-conditioning system;

[0059] The time lag value identification module of the central air conditioning system optimizes and identifies the time lag value of the central air conditioning system through data fitting and simulated annealing algorithm;

[0060] A time-delay stable interval prediction module for a central air-conditioning system, based on the time-delay value, applies the time-delay state space model of the central air-conditioning system to predict the time-delay stable interval of the central air-conditioning system;

[0061] The central air-conditioning system operation control module controls the operation of the central air-conditioning system according to the time-delay stable interval and in combination with fuzzy PID adaptive control.

[0062] Preferably, the time-lag value identification module of the central air-conditioning system includes a data fitting module and a minimum time-lag value search module, wherein the data fitting module is used to convert the time-lag problem into an optimization problem under constraints, and the minimum time-lag value search module searches for the minimum time-lag value under constraints through a simulated annealing algorithm.

[0063] Finally, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements an adaptive control method for a central air-conditioning system taking into account a time lag effect.

[0064] The present invention provides an adaptive control method and system for a central air-conditioning system taking into account the time lag effect. First, a time lag state space model based on energy balance is constructed, which can accurately describe the time lag phenomenon in the central air-conditioning system, thereby ensuring the stability of the system under different operating conditions. Secondly, a method combining data fitting and simulated annealing algorithm is adopted to accurately identify the time lag value in the system, avoiding the response lag problem caused by the time lag effect, thereby ensuring the stability of the system under different operating conditions. In addition, by designing a fuzzy PID control scheme that adapts to the time lag effect, the present invention optimizes the adaptive control capability of the central air-conditioning system, which not only improves the response speed of the system, but also reduces energy consumption, and significantly improves the overall performance of the system.

[0065] The present invention not only improves the control accuracy and energy efficiency of the central air-conditioning system during the demand response process, but also provides important guarantees for building energy conservation and stable operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A flow chart of the adaptive control method for a central air-conditioning system considering the time lag effect provided by the present invention;

[0067] Figure 2 A schematic diagram of a modeling process of a time-delay state space method in an embodiment disclosed in the present invention;

[0068] Figure 3 It is a schematic diagram of trend changes in the embodiments disclosed in the present invention;

[0069] Figure 4 A schematic diagram of a simulated annealing process in an embodiment disclosed in the present invention;

[0070] Figure 5 A schematic diagram of the fuzzy control structure in the disclosed embodiment of the present invention;

[0071] Figure 6 It is a schematic diagram of the structure of the fuzzy PID controller in the disclosed embodiment of the present invention. DETAILED DESCRIPTION

[0072] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0073] First, the present invention provides an adaptive control method for a central air-conditioning system taking into account the time lag effect, which improves the control accuracy, stability and energy efficiency of the system by accurately identifying and modeling the time lag effect, including:

[0074] S1: Construct a time-delay state space model of the central air-conditioning system based on energy balance to describe the time-delay phenomenon of the central air-conditioning system;

[0075] Time-delay state-space model: This invention uses an important branch of the state-space model, the time-delay state-space model, which regards the central air-conditioning system as a system with input time delay, and its expression is

[0076]

[0077] Where x(t) is the n-dimensional state vector x(t) = [x1(t), x2(t), ..., x n (t)] T , which may include state parameters such as temperature, pressure, flow rate, etc.; u(t) is an n-dimensional input vector u(t) = [u1(t),u2(t),...,u n (t)] T , which may include inlet temperature, inlet pressure, inlet flow, control opening, etc.; y(t) is the m-dimensional output vector y(t) = [y1(t), y2(t), ..., y n (t)] T , which may include state parameters such as outlet temperature, outlet pressure, outlet flow, and opening output; A and B are the coefficient matrix and input matrix respectively, C and D are the output matrix and direct transmission matrix respectively. The determination of the matrix value needs to be derived through the energy balance relationship of the modeling system or equipment. The specified model construction can be achieved through different coefficient matrices. It can be seen that the state vector reflects the influence of the input variables on the internal state of the system, and the output vector reflects the influence of the state variables on the output variables. Therefore, the use of the time-delay state space description method to establish the central air-conditioning distribution system component model can relatively clearly reflect the dynamic relationship between the input variables, state variables and output variables, and can make a more intuitive analysis of the structure of the constructed model.

[0078] The modeling of each link of the central air-conditioning system refers to the thermodynamic mathematical expressions such as the energy balance equation, and has a certain degree of unity, except that there are differences in the selection of parameter values ​​in the modeling link. Therefore, when using the time-delay state space method to model the central air-conditioning system, its basic modeling process is as follows: Figure 2 The specific design steps are as follows:

[0079] (1) Determine the basic equilibrium equation

[0080] As an important branch of state-space methods, the accuracy of time-delay state-space models is highly dependent on the comprehensiveness of the equilibrium equations in the modeling process. This modeling method requires a detailed description of all aspects of the system, among which the most critical are two types of equilibrium equations: mass balance and energy balance. The mass balance equation focuses on the flow of inlet and outlet materials in the modeling process to ensure the conservation of materials in the system. The energy balance equation is more complex and usually includes two aspects: heat balance and humidity balance. According to the specific modeling requirements and system characteristics, researchers can selectively apply these equilibrium equations. It is worth noting that the more equilibrium equations are included, the more accurately the final constructed model can reflect the behavior of the actual system. In most cases, these equilibrium equations follow the basic laws of thermodynamics, that is, the change in thermodynamic state is the sum of the energy absorbed or released by the medium and the energy exchanged to the outside world:

[0081]

[0082] Where T is temperature, °C; t is time, s; c is mass specific heat capacity, J / (kg·°C); a is heat transfer coefficient, W / (m 2 ·℃); ΔT1 and ΔT2 are the temperature differences after heat absorption and heat release and heat exchange, respectively, ℃.

[0083] (2) Convert to incremental form

[0084] Linearizing the basic equilibrium equations is the core goal of "converting to incremental form", and this process is achieved through two key steps. First, the basic variables in the equations (such as temperature, flow, and humidity) are restated as the sum of steady-state initial values ​​and small increments. Secondly, the coefficients in the equations (such as mass specific heat, heat transfer coefficient, and thermal conductivity) are linearized using the first-order Taylor expansion. By executing these steps, the original equations are successfully converted to incremental form, laying the foundation for subsequent analysis and calculations. It is worth noting that this method is widely used in mathematical modeling in fields such as thermodynamics and fluid mechanics, providing a powerful tool for simplified analysis of complex systems, allowing researchers to more effectively handle and analyze related problems.

[0085] (3) Determine variables

[0086] When constructing a time-delay state-space model, the selection of variables is a key step, and this process is closely related to the import and export conditions of the system. Generally speaking, the variables in the model can be divided into three categories: input variables, output variables, and state variables. For air-conditioning systems, the parameters at the inlet of the equipment are usually specified as input variables, while the export parameters are regarded as output variables. This configuration reflects the physical characteristics and operating logic of the system. As for the state variables, their selection has a certain flexibility and can be determined according to the specific modeling requirements. It is worth noting that the state variable can be a parameter at the inlet or a parameter at the outlet, depending on the specific requirements of the model and the characteristics of the system.

[0087] (4) Matrix representation

[0088] The formation of the coefficient matrix is ​​a key step in the process of constructing a time-delay state-space model. This process begins with the analysis of the increment and variable type, followed by the extraction of relevant quantities from the increment representation. Subsequently, the cumulative form is converted into a matrix product, and finally the multiple equilibrium equations are integrated into a unified matrix form. In this matrix structure, each element corresponds to a coefficient in a specific equilibrium equation. It is worth noting that the values ​​in the matrix are not fixed, but will be dynamically adjusted with changes in external conditions (such as environmental parameters such as indoor and outdoor temperature). Therefore, the resulting matrix actually reflects the characteristics of the system in a specific state.

[0089] (5)Convert to standard type

[0090] After completing the matrix construction, the next step is to convert the model into a standard format. This step involves writing the normalized state space equations and output equations using the previously determined matrices. This standardized representation not only simplifies the mathematical description of the model, but also lays the foundation for subsequent analysis and control design.

[0091] S2: Identify the time lag value of the central air conditioning system through data fitting and simulated annealing algorithm optimization;

[0092] Optionally, the time lag value identification method of step S2 is:

[0093] Since the data collected by the central air-conditioning system has a long period and a large time span between data, some characteristic information may be obscured, making it difficult to intuitively determine the time lag value generated during the system state change process. In order to accurately identify these time lag values, the change curves of two correlated state parameters (such as P1 and P2) in the system are first obtained through data fitting methods, and then compared.

[0094] Assume that the change curves of two correlated state parameters (P1, P2) in the air-conditioning system obtained by data fitting are as follows: Figure 3 shown.

[0095] Among them, device 2 is located after device 1 and is controlled by the operating data of device 1. After the demand for cooling and heating loads changes, the system will change from one steady state to another. At this time, taking P2 as a reference, within a certain time range, assuming that P1 is shifted to the right, data characteristics similar to the change trend of P2 will be obtained. At this time, the expression of P1 becomes P1(t-τ), and τ is the actual time lag of the system. The purpose of time lag identification is to obtain the time lag value between different state parameters. Considering the dimensional differences between the parameters, the data needs to be normalized, and then the corresponding change function is obtained by fitting. The minimum-maximum normalization method is used to map the data to the [0,1] interval. The formula is as follows:

[0096]

[0097] In the formula, max is the maximum value of the sample data, min is the minimum value of the sample data, and P is the conversion data. If the parameter change curve within a certain interval is obtained by fitting, the derivative of this function curve represents the rate of change of the parameter, and then the change trend between the parameters can be obtained. g1 (t-τ) and P g2 The mean of the absolute value of the derivative difference of (t) integrated over the range (t1, t2) is defined as P g1 With P g2 The trend of the time lag between is expressed as:

[0098]

[0099] Then, the time lag is comprehensively considered through the changing trend within the same daily sampling period. P1,P2 The smaller it is, the more consistent the trends of P1 and P2 are, and the smaller the hysteresis value is. When the P2 state parameter changes for the first time, the hysteresis is considered to end, and the actual hysteresis is t P1,P2 The value corresponding to the minimum value is

[0100] τ=τ min =argmint P1,P2 (5)

[0101] The corresponding time-lag series is

[0102] τ=[τ1,τ2,...,τ N ] (6)

[0103] Among them, as the state reference time point, τ N =0, indicating that the time lag of the air outlet variable at the last moment is 0, that is, the zero time lag reference variable; τ1~τ N-1In turn, they represent the time lag between the zero-time-lag reference variable and other corresponding groups of key state variables. A trend similarity matrix between multiple groups of data is constructed, and the expression is:

[0104]

[0105] In the formula, i = 1, 2, ..., N-1. tp reflects the trend consistency among multiple variables, and each element in the corresponding trend similarity vector tp is very small. Since it is a trend matrix, its size can be represented by a specific value. At this time, the concept of norm in numerical analysis is needed. The L2 norm is used to quantify the trend vector, and the minimum value of the L2 norm ||tp||2 of tp is obtained as the time lag value of the link described by the established state association data matrix. The expression is:

[0106] γ i =min(‖tp‖2) (8)

[0107] The τ obtained at this time may be a suboptimal solution, and further optimization calculations are needed to convert the time-delay problem into an optimization problem under constraints.

[0108]

[0109] In order to search for the minimum hysteresis value under the constraints, a simulated annealing algorithm is selected to find the optimal solution within a certain interval. The specific process of the algorithm is as follows: Figure 4 As shown. This algorithm refers to the annealing process of solids, that is, the process of slow cooling of solids after being heated. It uses the disordered state of particles inside the solids when heated and the tendency of particles to become ordered when slowly cooled to complete the optimization problem. After any random temperature interference, the algorithm calculates the new objective function and determines whether the modification is accepted, and then records it. When the temperature interference changes, the new objective function is calculated here. When the calculation is stopped, all objective function values ​​are compared and the optimal value is selected for output. The control parameter in the algorithm is temperature T, the objective function is internal energy E, and the state node at temperature T is x i When the control parameter T decreases, the internal energy E of the objective function will also decrease until the temperature reaches the lowest energy state. At this time, the global minimum value searched is the optimal value of the data trend similarity matrix.

[0110] The model compensates the data within the range according to the sampling time. According to the identification of the time lag, within the sampling period, for example, when the time lag identification values ​​are 1 to 5 minutes, the data is supplemented, as shown below:

[0111]

[0112] If τ is greater than the sampling period (5 min), the compensation result appears in the next period.

[0113] This time-lag value identification method that combines data fitting with simulated annealing algorithm can accurately identify the time-lag effect in the central air-conditioning system, and ensure the global optimality of the identification results through further optimization, thereby improving the control accuracy and operation efficiency of the system.

[0114] S3: Based on the identified time lag values, the time lag state space method is applied to establish a linear model of energy balance, predict the time lag stability interval of the system, and ensure the stability of the system under different operating conditions;

[0115] Here, by applying the time-delay state space method, a linear energy balance model of the central air-conditioning system is established to accurately describe the time-delay phenomenon in the system and further predict the time-delay stability interval of the system under different working conditions. The time-delay value in the central air-conditioning system is a dynamically changing parameter. Therefore, in order to ensure the stability of the system under various operating conditions, it is necessary to accurately determine the stability interval of the time-delay.

[0116] In the process of model establishment, the integral inequality method is used to establish the stability criterion related to the time lag range of the model. Since the time lag in the air-conditioning system is a changing value, the changing speed of the time lag value is set to u, the lower bound of the time lag stability is τ1, and the upper bound is τ2.

[0117] e i is a block coordinate matrix (i=1,...,9), for example,

[0118]

[0119] in,

[0120] The following theorem gives the relevant conditions for the range of time delays that satisfy the first case. For a given scalar 0<τ1<τ2, 0≤μ<1, if there exists a matrix Q of suitable dimension i >0(i=1,...,5), Z j >0(j=1,...,4), R1>0, R2>0, P=[P ij ] 5×5 >0, the system or model is asymptotically stable if the following linear matrix inequality holds:

[0121]

[0122] That

[0123]

[0124] This method can effectively predict the time-delay stability range of the system under different operating conditions. Specifically, the model can determine the upper and lower limits of the time-delay of the central air-conditioning system under various operating conditions, thereby ensuring that the system is always in a stable state under these conditions.

[0125] S4: Combined with fuzzy PID control, an adaptive control scheme for the central air-conditioning system is designed to adapt to the time lag effect and improve the system response speed and efficiency.

[0126] Here, fuzzy-PID control overcomes the shortcomings of a single control method in dealing with complex working conditions by integrating fuzzy control theory with traditional PID control algorithms, and realizes a more stable and efficient control system. In central air-conditioning systems, due to the existence of time lag effects, system response may be delayed or overshoot. In order to effectively solve these problems, the present invention designs an adaptive control scheme that adapts to time lag effects.

[0127] In the inner control design, fuzzy control is responsible for handling the qualitative, fuzzy and uncertain state parameters that are widely present in the air-conditioning system. For example, in the control process of parameters such as hot and cold temperatures, flow rates, and operating time, fuzzy control can convert these fuzzy areas (such as neither hot nor cold, neither big nor small, neither long nor short) into fuzzy subsets between 0 and 1 through membership functions to describe the current state of the system. Next, fuzzy reasoning determines the appropriate control output based on the real-time state information of the system and the fuzzy reasoning rules in the database, and accurately processes the fuzzified control value, which directly acts on the actuator of the air-conditioning system, thereby realizing dynamic adjustment of the system. Figure 5 The structure of the fuzzy control system is shown.

[0128] The PID link in Fuzzy-PID is to solve the problems of reduced control accuracy and poor dynamic quality of the controlled link caused by fuzzy processing of system information. The Fuzzy-PID controller will determine the composition and parameter settings of PID through fuzzy reasoning. The control structure is as follows: Figure 6 As shown. Among them, the algorithm expression of PID controller is

[0129]

[0130] In the formula, k p is the proportionality coefficient; T i is the integration time constant; T d is the differential time constant; e(t) is the control deviation e(t) = r(t)-y(t).

[0131] Through this fuzzy PID control scheme, the present invention can adjust the system control strategy in real time during the operation of the air-conditioning system, adapt to different operating environments and working conditions, effectively improve the response speed and control accuracy of the system, and significantly improve the overall performance and energy efficiency of the central air-conditioning system, especially in complex demand response scenarios, to ensure stable and efficient operation of the system

[0132] This method can not only accurately describe the time lag phenomenon in the system, but also improve the control accuracy and energy efficiency of the system. This method will help solve the control lag problem of the central air-conditioning system in the demand response process and improve the overall operation efficiency of the system. It has important theoretical significance and practical application value.

[0133] Secondly, the present invention also provides an adaptive control system for a central air conditioning system taking into account the time lag effect, comprising:

[0134] A module for building a time-delay state-space model of a central air-conditioning system based on energy balance, which is used to describe the time-delay phenomenon of the central air-conditioning system;

[0135] The time lag value identification module of the central air conditioning system optimizes and identifies the time lag value of the central air conditioning system through data fitting and simulated annealing algorithm;

[0136] A time-delay stable interval prediction module for a central air-conditioning system, based on the time-delay value, applies the time-delay state space model of the central air-conditioning system to predict the time-delay stable interval of the central air-conditioning system;

[0137] The central air-conditioning system operation control module controls the operation of the central air-conditioning system according to the time-delay stable interval and in combination with fuzzy PID adaptive control.

[0138] The time lag value identification module of the central air conditioning system includes a data fitting module and a minimum time lag value search module, wherein the data fitting module is used to convert the time lag problem into an optimization problem under constraints, and the minimum time lag value search module searches for the minimum time lag value under constraints through a simulated annealing algorithm. Finally, this embodiment also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the above-mentioned central air conditioning system adaptive control method considering the time lag effect is implemented.

[0139] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An adaptive control method for a central air conditioning system considering time lag effect, characterized in that: include: Construct a time-delay state-space model of central air-conditioning system based on energy balance to describe the time-delay phenomenon of central air-conditioning system; Through data fitting and simulated annealing algorithm optimization, the time lag value of the central air conditioning system is identified; Based on the time lag value, applying the central air conditioning system time lag state space model to predict the time lag stability interval of the central air conditioning system; Based on the time-delay stable interval, the operation of the central air-conditioning system is controlled by combining fuzzy PID adaptive control; The time-delay state space model of the central air conditioning system is expressed as follows: Where x(t) is the n-dimensional state vector x(t) = [x1(t), x2(t), ..., x n (t)] T , including the state parameters of temperature, pressure and flow rate; u(t) is the n-dimensional input vector u(t)=[u1(t),u2(t),...,u n (t)] T , including inlet temperature, inlet pressure, inlet flow, and control opening; y(t) is the m-dimensional output vector y(t) = [y1(t), y2(t), ..., y n (t)] T , including the state parameters of outlet temperature, outlet pressure, outlet flow, and opening output; A and B are the coefficient matrix and input matrix respectively, and C and D are the output matrix and direct transmission matrix respectively.

2. The adaptive control method of central air conditioning system considering time lag effect according to claim 1, characterized in that: The construction of a time-delay state space model of a central air-conditioning system based on energy balance includes: 1) Determine the basic equilibrium equation: The change in thermodynamic state is the sum of the energy absorbed or released by the medium and the energy exchanged with the outside world: Where T is temperature, °C; t is time, s; c is mass specific heat capacity, J / (kg·°C); a is heat transfer coefficient, W / (m 2 ·℃); ΔT1 and ΔT2 are the temperature differences after heat absorption and heat exchange, respectively, ℃; 2) Linearize the basic equilibrium equation and convert it into incremental form: Re-express the basic variables in the equation as the sum of the steady-state initial value and a small increment; linearize the coefficients in the equation using the first-order Taylor expansion; 3) Determine variables: specify the inlet parameters of the central air-conditioning system as input variables, and the outlet parameters of the central air-conditioning system as output variables; determine the state variables according to the modeling requirements; 4) Forming the coefficient matrix: Analyze the increment and variable type and extract relevant quantities from the increment representation; Convert the cumulative form into a matrix product, and finally integrate multiple equilibrium equations into a unified matrix form; 5) Convert to standard form and use the coefficient matrix to write the normalized state space equations and output equations.

3. The adaptive control method of central air conditioning system considering time lag effect according to claim 1, characterized in that: Through the data fitting method, the time lag problem is converted into an optimization problem under constraints; through the simulated annealing algorithm, the minimum time lag value under the constraints is searched out.

4. The adaptive control method of central air conditioning system considering time lag effect according to claim 3 is characterized in that: The method of data fitting is used to convert the time lag problem into an optimization problem under constraints, including: Obtain two correlated state parameter change curves in the central air-conditioning system and compare them; use the minimum-maximum normalization method to map the data to the [0,1] interval, and the formula is as follows: In the formula, max is the maximum value of the sample data, min is the minimum value of the sample data, and P is the conversion data; g1 (t-τ) and P g2 The mean of the absolute value of the derivative difference of (t) integrated over the range (t1, t2) is defined as P g1 With P g2 The trend of the time lag between τ and τ is the actual time lag of the system; the expression is: The size of the time lag is comprehensively considered through the changing trend within the same daily sampling period; The smaller it is, the more consistent the trends of P1 and P2 are, and the smaller the hysteresis value is. When the P2 state parameter changes for the first time, the hysteresis is considered to end, and the actual hysteresis is The value corresponding to the minimum value is The corresponding time-lag series is τ=[τ1,τ2,...,τ N ] (6) Among them, as the state reference time point, τ N =0, indicating that the time lag of the air outlet variable at the last moment is 0, that is, the zero time lag reference variable; τ1~τ N-1 In turn, they represent the time lag between the zero-time-lag reference variable and other corresponding groups of key state variables; constructing the trend similarity matrix between multiple groups of data, the expression is: In the formula, i = 1, 2, ..., N-1; tp reflects the trend consistency among multiple variables; The L2 norm is used to quantify the trend vector, and the minimum value of the L2 norm of tp||tp||2 is obtained as the time lag value of the link described by the established state association data matrix. The expression is: c i =min(‖tp‖2) (8) The τ obtained at this time may be a suboptimal solution. Further optimization calculations are performed to convert the time-delay problem into an optimization problem under constraints.

5. The adaptive control method of central air conditioning system considering time lag effect according to claim 3 is characterized in that: The simulated annealing algorithm is used to search for the minimum hysteresis value under the constraint conditions, including: the control parameter is temperature T, the objective function is internal energy E, and the state node at temperature T is x i ; When the control parameter T decreases, the internal energy E of the objective function will also decrease until the temperature reaches the lowest energy state. At this time, the global minimum value searched is the optimal value of the data trend similarity matrix.

6. The adaptive control method of central air conditioning system considering time lag effect according to claim 1, characterized in that: The method of predicting the time-delay stability interval of the central air-conditioning system by applying the time-delay state space model of the central air-conditioning system based on the time-delay value includes: The integral inequality method is used to establish the stability criterion related to the time-delay range of the time-delay state space model prediction of the central air-conditioning system. Let the change speed of the time-delay value be u, the lower bound of the time-delay stability be τ1, and the upper bound be τ2. e i is the block coordinate matrix (i=1,...,9), For a given scalar 0<τ1<τ2, 0≤μ<1, if there exists a matrix Q of suitable dimension i >0(i=1,...,5), Z j >0(j=1,...,4), R1>0, R2>0, P=[P ij ] 5×5 >0, the following linear matrix inequality holds, and the model is asymptotically stable: in, 7. The adaptive control method of central air conditioning system considering time lag effect according to claim 1, characterized in that: The method of controlling the operation of the central air-conditioning system based on the time-delay stable interval and combining fuzzy PID adaptive control includes: Fuzzy PID adaptive control converts the fuzzy regions of related parameters such as hot and cold temperatures, flow rate, and running time into fuzzy subsets between 0 and 1 through membership functions to describe the current state of the system. Fuzzy reasoning controls the output based on the real-time status information of the central air-conditioning system and the fuzzy reasoning rules in the database, and processes the fuzzy control value precisely, directly acting on the actuator of the air-conditioning system, thereby realizing dynamic adjustment of the system.

8. The adaptive control method of central air conditioning system considering time lag effect according to claim 7, characterized in that: The fuzzy-PID controller is used in the fuzzy PID adaptive control, wherein the algorithm expression of the PID controller is: In the formula, k p is the proportionality coefficient; T i is the integration time constant; T d is the differential time constant; e(t) is the control deviation e(t) = r(t)-y(t).

9. An adaptive control system for a central air conditioning system considering time lag effect, characterized in that: include: A module for building a time-delay state-space model of a central air-conditioning system based on energy balance, which is used to describe the time-delay phenomenon of the central air-conditioning system; The time lag value identification module of the central air conditioning system optimizes and identifies the time lag value of the central air conditioning system through data fitting and simulated annealing algorithm; A time-delay stable interval prediction module for a central air-conditioning system, based on the time-delay value, applies the time-delay state space model of the central air-conditioning system to predict the time-delay stable interval of the central air-conditioning system; The central air-conditioning system operation control module controls the operation of the central air-conditioning system according to the time-delay stable interval and in combination with fuzzy PID adaptive control.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the adaptive control method of a central air-conditioning system considering the time lag effect described in any one of claims 1 to 8 is implemented.

Images