An operating parameter optimization control method and system for an air condenser

Abstract

The present invention relates to the field of optimization control technology, and specifically to an operation parameter optimization control method and system for an air condenser. First, environmental data is acquired and processed to obtain an adaptive environmental monitoring value, and a non-uniform frequency conversion strategy is constructed according to the adaptive environmental monitoring value to control the optimization frequency of the operation parameters of the air condenser. Then, the structural characteristics and operation characteristics of the air condenser are acquired, and a first parameter vector to be optimized is obtained according to the principal component analysis method. The first parameter vector to be optimized is acquired and sensitivity analysis is performed to obtain a second parameter vector to be optimized. Subsequently, an optimized parameter prediction model predicts the parameter value vector of the second parameter vector to be optimized, and a distributed parameter optimization control model sets local agents for the parameters to be optimized. The parameter value vector of the second parameter vector to be optimized is acquired and parameter optimization control is performed through a reinforcement learning algorithm. The present invention improves the parameter optimization speed and stability and reduces the energy consumption of the air condenser through the optimization control of the operation parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of optimization control, and specifically provides an operation parameter optimization control method and system for an air condenser. Background Art

[0002] Currently, the operation parameter optimization control methods for air condensers mainly include using fixed set values or simple feedback control, simple PID control, and timed start-stop control. For example, frequently starting and stopping the condenser fan at certain time intervals can easily lead to poor control effects and increased equipment wear.

[0003] Traditional control algorithms are difficult to achieve fast and stable optimization control of operation parameters under different working conditions. Therefore, it is necessary to use adaptive control methods such as deep reinforcement learning and fuzzy control, and combine with the usage of the equipment itself to dynamically adjust the optimization frequency of the operation parameters of the air condenser, so as to achieve precise and stable optimization control of the operation parameters.

[0004] Therefore, an operation parameter optimization control method and system for an air condenser are proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide an operation parameter optimization control method and system for an air condenser. First, obtain environmental data and process it to obtain an adaptive environmental monitoring value, and construct a non-uniform frequency conversion strategy based on the adaptive environmental monitoring value to control the optimization frequency of the operation parameters of the air condenser; then obtain the structural characteristics and operation characteristics of the air condenser and obtain the first parameter vector to be optimized according to the principal component analysis method, obtain the first parameter vector to be optimized and perform sensitivity analysis to obtain the second parameter vector to be optimized; subsequently, the optimized parameter prediction model predicts the parameter value vector of the second parameter vector to be optimized, the distributed parameter optimization control model sets the local intelligent agent of the parameter to be optimized, obtains the parameter value vector of the second parameter vector to be optimized, and performs parameter optimization control through the reinforcement learning algorithm.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] An operation parameter optimization control method for an air condenser, comprising:

[0008] Obtain environmental data and process it to obtain an adaptive environmental monitoring value, and construct a non-uniform frequency conversion strategy based on the adaptive environmental monitoring value to control the optimization frequency of the operation parameters of the air condenser;

[0009] Further, the adaptive environmental monitoring value includes a thermal stress term, a radiation term, and a time-varying correction term, and the calculation formula of the adaptive environmental monitoring value is:

[0010] ;

[0011] Among them, represents the adaptive environmental monitoring value, and represents the adaptive weight coefficient, represents the temperature, represents the temperature nonlinear coefficient, represents the sign function, represents the preset temperature, represents the humidity coupling coefficient, represents the humidity, represents the saturated humidity, represents the wind speed attenuation coefficient, represents the square root of the wind speed, represents the radiant heat intensity, and represents the pressure modulation parameter, represents the natural base, represents the atmospheric pressure, represents the standard atmospheric pressure, and represents the historical trend sensitivity coefficient, represents the hyperbolic tangent function, represents the temperature change trend.

[0012] Furthermore, the non-uniform frequency conversion strategy includes a transient response term, a memory decay integral term, and a nonlinear compensation term; the calculation formula of the non-uniform frequency conversion strategy is:

[0013] ;

[0014] Among them, represents the non-uniform frequency conversion strategy, and represent the transient response term coefficient and the memory decay integral term coefficient, represents the derivative of the adaptive environmental monitoring value, represents the Gaussian kernel convolution, represents the bandwidth adaptively adjusted according to the signal-to-noise ratio, represents the adaptive environmental monitoring value function, represents the memory decay coefficient, represents the time, represents the integration variable, represents the nonlinear compensation term coefficient, represents the sign function, represents the gradient operator, represents the Laplacian term.

[0015] Obtain the structural characteristics and operating characteristics of the air condenser and obtain the first parameter vector to be optimized according to the principal component analysis method; obtain the first parameter vector to be optimized and perform sensitivity analysis to obtain the second parameter vector to be optimized.

[0016] Furthermore, obtain the parameter values in the first parameter vector to be optimized and calculate the sensitivity index of the parameter values through sensitivity analysis to obtain a set of sensitivity indices; obtain the elements in the set of sensitivity indices, sort them, and calculate the cumulative contribution degree. According to the parameter to be optimized corresponding to the sensitivity index when the cumulative contribution degree exceeds the threshold, construct the second parameter vector to be optimized.

[0017] The optimization parameter prediction model obtains all the parameters and characteristics of the air condenser and makes predictions to obtain the parameter value vector of the second parameter vector to be optimized.

[0018] Furthermore, the steps of constructing the optimization parameter prediction model include:

[0019] Step S1: Establish an objective comprehensive evaluation mechanism to obtain multi-objective scoring labels. The objective comprehensive evaluation mechanism includes a heat transfer objective, an energy-saving objective, and an equipment status objective; set initial weights through the AHP analytic hierarchy process and use an adaptive fuzzy inference system to adjust the weight values in real time.

[0020] Step S2: Obtain historical environmental data, operating parameters, and equipment status data and capture the equipment response characteristics in the time series through a multi-layer LSTM network and an attention enhancement mechanism to obtain equipment characteristics.

[0021] Step S3: Train the XGBoost regression model.

[0022] Step S4: Obtain the trained XGBoost regression model and make predictions. The model inputs include environmental characteristics, equipment characteristics, historical parameters, and multi-objective scoring labels; the model outputs include the parameter value vector of the second parameter vector to be optimized.

[0023] The distributed parameter optimization control model includes local agents for setting parameters to be optimized. The local agents obtain the parameter value vector of the second parameter vector to be optimized and perform parameter optimization control through a reinforcement learning algorithm.

[0024] Furthermore, the steps of constructing the distributed parameter optimization control model include: Step S10: Configure a local agent for each parameter and perform local control using a deep reinforcement learning algorithm; the state space of the local agent includes current environmental data, equipment operating status, and status information of the parameter value vector; the action space includes the adjustment range of the corresponding parameter.

[0025] Step S20: Construct a reward function for the local agent based on the heat exchange efficiency, energy consumption, and parameter variation range;

[0026] Step S30: Train according to the deep reinforcement learning algorithm to obtain the trained local agent;

[0027] Step S40: Obtain the parameter value vectors of the trained local agent and the second parameter vector to be optimized. According to the optimized parameter values and the current actual parameter values, if the deviation exceeds the set threshold, adopt a progressive adjustment strategy;

[0028] Step S50: Obtain the cumulative reward of the local agent. After reaching the set convergence condition of the objective function or reaching the maximum number of iterations, stop the iteration, and the control execution of the local agent ends.

[0029] The present invention also provides an operating parameter optimization control system for an air condenser, including:

[0030] A control frequency optimization module, including an adaptive environmental monitoring value calculation unit and a non-uniform frequency conversion strategy control unit;

[0031] A parameter to be optimized selection module, configured to obtain the structural characteristics and operating characteristics of the air condenser and obtain a first parameter vector to be optimized according to the principal component analysis method; obtain the first parameter vector to be optimized and perform sensitivity analysis to obtain a second parameter vector to be optimized;

[0032] An optimized parameter prediction module, configured to obtain all parameters and characteristics and perform prediction through multi-objective modeling, feature extraction, and an XGBoost regression model to obtain the parameter value vector of the second parameter vector to be optimized;

[0033] The distributed control module includes a distributed parameter optimization control model, configured to set local agents for the parameters to be optimized, obtain the parameter value vector of the second parameter vector to be optimized, and perform parameter optimization control through a reinforcement learning algorithm.

[0034] Compared with the prior art, the beneficial effects of the present invention are:

[0035] 1. By acquiring and processing real-time environmental data and constructing an adaptive environmental monitoring value including a thermal stress term, a radiation term, and a time-varying correction term, the system can accurately reflect the dynamic changes of multi-dimensional environmental factors such as temperature, humidity, wind speed, radiation intensity, and atmospheric pressure; the non-uniform frequency conversion strategy constructed using the adaptive environmental monitoring value combines three parts: a transient response term, a memory decay integral term, and a non-linear compensation term, which can take into account both the short-term rapid response and the long-term historical trend influence of the system, so as to ensure that the optimization frequency of the operating parameters can quickly respond to sudden changes.

[0036] 2. By calculating the principal component analysis method and sensitivity index, the parameters with the greatest impact on heat transfer, energy conservation, and equipment status are selected to form the second parameter vector to be optimized, which can ensure that the optimization control focuses on the most sensitive and critical parameters, thereby improving the overall control effect and robustness. The target comprehensive evaluation mechanism is adopted to unify the modeling of multiple key objectives such as heat transfer, energy conservation, and equipment status, construct a multi-objective scoring label, and then use the XGBoost regression model for prediction, which can integrate multi-dimensional features and achieve high-precision prediction of the key parameter value vector.

[0037] 3. By constructing a reward function based on heat transfer efficiency, energy consumption, and parameter change range, it is possible to pursue the best energy efficiency and heat exchange performance on the premise of ensuring the safe and stable operation of the equipment. This reward mechanism ensures that the control strategy achieves global optimization. The distributed parameter optimization control model can achieve the balanced optimization of multiple objectives such as heat transfer and energy consumption while ensuring system safety and equipment protection, and improve the overall control efficiency. Brief Description of the Drawings

[0038] Figure 1 It is a flowchart of an operation parameter optimization control method for an air condenser provided by an embodiment of the present invention;

[0039] Figure 2 It is a flowchart of constructing an optimized parameter prediction model provided by an embodiment of the present invention;

[0040] Figure 3 It is a schematic structural diagram of an operation parameter optimization control system for an air condenser provided by an embodiment of the present invention. Detailed Embodiments

[0041] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0042] Embodiment 1

[0043] In order to improve the working efficiency of the cooling system and energy efficiency, a certain data center introduces an operation parameter optimization control method for an air condenser provided by the present invention. The method flow is as Figure 1 shown, and the specific implementation is as follows:

[0044] Obtain environmental data and process it to obtain an adaptive environmental monitoring value, and construct a non-uniform frequency conversion strategy based on the adaptive environmental monitoring value to control the optimization frequency of the operation parameters of the air condenser;

[0045] Further, the adaptive environmental monitoring value includes a thermal stress term, a radiation term, and a time-varying correction term, corresponding to the first, second, and third terms in the formula. The calculation formula for the adaptive environmental monitoring value is:

[0046] ;

[0047] where, represents the adaptive environmental monitoring value, and represent the adaptive weight coefficients, represents the temperature, represents the temperature nonlinear coefficient, represents the sign function, represents the preset temperature, represents the humidity coupling coefficient, represents the humidity, represents the saturated humidity, represents the wind speed attenuation coefficient, represents the square root of the wind speed, represents the radiant heat intensity, and represent the pressure modulation parameters, represents the natural logarithm base, represents the atmospheric pressure, represents the standard atmospheric pressure, and represent the historical trend sensitivity coefficients, represents the hyperbolic tangent function, represents the temperature change trend;

[0048] Further, the calculation formula for the temperature change trend is:

[0049] ;

[0050] where, represents the time window, represents the time, represents the temperature function;

[0051] Further, the adaptive weights and are dynamically updated through Kalman filtering. Table 1 shows the adaptive environmental monitoring values at different time nodes.

[0052] Table 1. Adaptive Environmental Monitoring Values

[0053]

[0054] The adaptive environmental monitoring value synthesizes various factors such as temperature, humidity, wind speed, radiant heat, and atmospheric pressure, and at the same time introduces parameters such as temperature non-linearity, humidity coupling, and historical trend sensitivity coefficient, so that the monitoring value can more accurately reflect the dynamic changes of the actual environment, helping to avoid the deviation that may be introduced by a single index and improving the accuracy of environmental monitoring.

[0055] Further, the non-uniform frequency conversion strategy includes a transient response term, a memory decay integral term, and a non-linear compensation term, corresponding to the first, second, and third terms in the formula. The calculation formula of the non-uniform frequency conversion strategy is:

[0056] ;

[0057] Among them, represents the non-uniform frequency conversion strategy, and represent the transient response term coefficient and the memory decay integral term coefficient, represents the derivative of the adaptive environmental monitoring value, represents the Gaussian kernel convolution, represents the bandwidth adaptively adjusted according to the signal-to-noise ratio, represents the adaptive environmental monitoring value function, represents the memory decay coefficient, represents time, represents the integration variable, represents the non-linear compensation term coefficient, represents the sign function, represents the gradient operator, represents the Laplace term;

[0058] Further, it is judged whether to optimize the operating parameters of the air condenser according to the non-uniform frequency conversion strategy.

[0059] This non-uniform frequency conversion strategy realizes a comprehensive response to environmental signals through three complementary parts, and can intelligently judge whether it is necessary to optimize the operating parameters of the air condenser. When the environment changes greatly, the optimization frequency automatically increases to quickly respond; while when the environment is stable, the optimization frequency is reduced to avoid frequent adjustment, thereby saving computing resources and reducing equipment wear.

[0060] Obtain the structural characteristics and operating characteristics of the air condenser and obtain the first parameter vector to be optimized according to the principal component analysis method; obtain the first parameter vector to be optimized and conduct a sensitivity analysis to obtain the second parameter vector to be optimized;

[0061] Further, obtain the parameter values in the first parameter vector to be optimized and calculate the sensitivity index of the parameter through sensitivity analysis to obtain a set of sensitivity indices; obtain the elements in the set of sensitivity indices, sort them, and calculate the cumulative contribution degree. Based on the parameter to be optimized corresponding to the sensitivity index when the cumulative contribution degree exceeds the threshold, construct a second parameter vector to be optimized.

[0062] Further, the calculation formula for the sensitivity index is:

[0063] ;

[0064] where represents the sensitivity index of the th parameter to be optimized in the first parameter vector to be optimized , represents the system response variable, represents the variance, represents the expectation;

[0065] The sensitivity index can measure the contribution of each parameter to the system response variable. By sorting and calculating the cumulative contribution degree, parameters with less influence on the response can be eliminated, thereby reducing the dimension of the parameters to be optimized, helping to reduce the computational complexity of the subsequent optimization control algorithm, and improving the control speed.

[0066] The optimized parameter prediction model obtains all the parameters and features and performs prediction through multi-objective modeling, feature extraction, and the XGBoost regression model to obtain the parameter value vector of the second parameter vector to be optimized;

[0067] Further, the steps for constructing the optimized parameter prediction model are as Figure 2 shown, including:

[0068] Step S1: Establish an objective comprehensive evaluation mechanism to obtain multi-objective scoring labels. The objective comprehensive evaluation mechanism includes a heat exchange objective, an energy-saving objective, and an equipment status objective; set initial weights through the AHP analytic hierarchy process and adjust the weight values in real time using adaptive fuzzy inference;

[0069] Further, the heat exchange objective , where represents the maximum heat exchange efficiency, represents the weight coefficient for reducing the condensation temperature ; the energy-saving objective , where represents the minimum operating power consumption, represents the energy-saving benefit ; the equipment status objective , where represents the degree of equipment wear, Represents the equipment maintenance cost The weight coefficient of;

[0070] Furthermore, the weight coefficient , And Is determined by the AHP (Analytic Hierarchy Process) and adaptive fuzzy inference.

[0071] Step S2: Obtain historical environment data, operating parameters, and equipment status data, and capture the equipment response characteristics in the time series through a multi-layer LSTM network and an attention enhancement mechanism to obtain equipment features;

[0072] Step S3: Train the XGBoost regression model. The training dataset includes multi-objective scoring labels and a dataset. The dataset includes environmental features, equipment features, and historical parameters;

[0073] Step S4: Obtain the trained XGBoost regression model and make predictions. The model inputs include environmental features, equipment features, historical parameters, and multi-objective scoring labels; the model outputs include the parameter value vector of the second parameter vector to be optimized.

[0074] Adopt a target comprehensive evaluation mechanism to obtain multi-objective scoring labels, unify the modeling of multiple key objectives such as heat exchange, energy conservation, and equipment status, so that the system does not favor a single objective during the optimization process, thereby achieving a balanced improvement in overall performance; set initial weights through the AHP, and use an adaptive fuzzy inference system to adjust the weight values in real time, which can flexibly correct the importance of objectives according to the actual operating conditions and environmental changes; use the XGBoost regression model for prediction, which can integrate multi-dimensional features and achieve high-precision prediction of the key parameter value vector.

[0075] The distributed parameter optimization control model obtains the parameter value vector of the second parameter vector to be optimized and performs parameter optimization control; the distributed parameter optimization control model includes local agents for setting parameters to be optimized and performs parameter optimization control through a reinforcement learning algorithm.

[0076] Furthermore, the steps for constructing the distributed parameter optimization control model include:

[0077] Step S10: Configure a local agent for each parameter and perform local control using a deep reinforcement learning algorithm; the state space of the local agent includes the current environmental data, equipment operating status, and status information of the parameter value vector; the action space includes the adjustment range of the corresponding parameter;

[0078] Step S20: Construct a reward function for the local agent according to the heat exchange efficiency, energy consumption, and parameter change range;

[0079] Furthermore, the formula for the reward function is:

[0080] ;

[0081] Among them, represents the reward function, represents the hyperbolic tangent function, represents the actual heat exchange efficiency, represents the reference heat exchange efficiency, represents the maximum heat exchange efficiency, represents the logarithmic function, represents the actual power, represents the rated power, represents the parameter change amount of adjacent time steps, represents the two-norm.

[0082] Step S30: Train the local agent according to the deep reinforcement learning algorithm to obtain the trained local agent;

[0083] Step S40: Obtain the parameter value vectors of the trained local agent and the second parameter vector to be optimized. According to the optimized parameter values and the current actual parameter values, if the deviation exceeds the set threshold, adopt a progressive adjustment strategy;

[0084] Furthermore, the progressive adjustment strategy means that if the deviation exceeds the set threshold, the deviation is segmented at multiple levels, and then the parameter optimization control is carried out step by step.

[0085] Step S50: Obtain the cumulative reward of the local agent. After reaching the set target function convergence condition or the maximum number of iterations, stop the iteration, and the control execution of the local agent ends. Table 2 shows the data when the local agent performs parameter optimization control.

[0086] Table 2. Operating parameter control data

[0087]

[0088] Each parameter is responsible for by a dedicated local agent, and its state space covers the current environmental data, equipment operating status and parameter value information, enabling each local agent to make fine adjustments according to the specific dynamics of the parameter, so as to achieve more accurate local control; by constructing a reward function based on heat exchange efficiency, energy consumption and parameter change range, it is possible to pursue the best energy efficiency and heat exchange performance while ensuring the safe and stable operation of the equipment, and achieve global optimization; when the cumulative reward reaches the set convergence condition or the iteration number limit, the system automatically stops adjusting, ensuring that the entire control process converges to the optimal state quickly and efficiently.

[0089] First, by collecting and processing environmental data in real time to obtain adaptive environmental monitoring values, the dynamic changes of the environment can be captured, thereby constructing a non-uniform frequency conversion strategy to flexibly adjust parameters and optimize the frequency to cope with external disturbances. Second, principal component analysis and sensitivity analysis are used to screen out the key parameters that have the greatest impact on the system response from numerous structural and operating characteristics, significantly reducing the dimension of the parameters to be optimized and improving the computational efficiency and accuracy of subsequent optimization. Subsequently, the target comprehensive evaluation mechanism combines feature extraction and the XGBoost regression model to accurately predict the optimal values of the key parameters and achieve a comprehensive balance of thermal efficiency, energy conservation, and equipment status. Finally, through the distributed parameter optimization control model, local intelligent agents optimize parameters based on the deep reinforcement learning algorithm to achieve real-time, robust, and safe optimization control of global parameters. Generally speaking, the method provided by the present invention greatly improves the response speed, energy efficiency, and stability of parameter optimization, and reduces energy consumption and equipment wear.

[0090] Embodiment 2

[0091] The present invention also provides an operating parameter optimization control system for an air condenser. The system structure is as Figure 3 shown, and the specific implementation is as follows:

[0092] The control frequency optimization module includes an adaptive environmental monitoring value calculation unit and a non-uniform frequency conversion strategy control unit;

[0093] Furthermore, environmental data is acquired and processed to obtain adaptive environmental monitoring values, and a non-uniform frequency conversion strategy is constructed based on the adaptive environmental monitoring values to control the optimized frequency of the operating parameters of the air condenser.

[0094] Furthermore, the adaptive environmental monitoring values include a thermal stress term, a radiation term, and a time-varying correction term. The calculation formula for the adaptive environmental monitoring values is:

[0095] ;

[0096] Where represents the adaptive environmental monitoring value, and represent the adaptive weight coefficients, represents the temperature, represents the temperature nonlinear coefficient, represents the sign function, represents the preset temperature, represents the humidity coupling coefficient, represents the humidity, represents the saturated humidity, represents the wind speed attenuation coefficient, represents the square root of the wind speed, represents the radiant heat intensity, and represents the pressure modulation parameter, represents the natural base, represents the atmospheric pressure, represents the standard atmospheric pressure, and represents the historical trend sensitivity coefficient, represents the hyperbolic tangent function, represents the temperature change trend.

[0097] Furthermore, the non-uniform frequency conversion strategy includes a transient response term, a memory decay integral term, and a non-linear compensation term; the calculation formula of the non-uniform frequency conversion strategy is:

[0098] ;

[0099] wherein, represents the non-uniform frequency conversion strategy, and represent the transient response term coefficient and the memory decay integral term coefficient, represents the derivative of the adaptive environmental monitoring value, represents the Gaussian kernel convolution, represents the bandwidth adaptively adjusted according to the signal-to-noise ratio, represents the adaptive environmental monitoring value function, represents the memory decay coefficient, represents time, represents the integration variable, represents the non-linear compensation term coefficient, represents the sign function, represents the gradient operator, represents the Laplacian term. As shown in Table 3, it is judged whether to perform parameter optimization control according to the non-uniform frequency conversion strategy at different time nodes, and the control accuracy rate is increased to 98.3%, the false trigger rate is reduced to 3.1%, and the response delay is controlled within 50 ms.

[0100] Table 3. Parameter Optimization Control Decisions at Partial Time Nodes

[0101]

[0102] The to-be-optimized parameter selection module is used to obtain the structural characteristics and operating characteristics of the air condenser and obtain the first to-be-optimized parameter vector according to the principal component analysis method; obtain the first to-be-optimized parameter vector and perform sensitivity analysis to obtain the second to-be-optimized parameter vector;

[0103] Further, obtain the parameter values in the first parameter vector to be optimized and calculate the sensitivity index of the parameter through sensitivity analysis to obtain a set of sensitivity indices; obtain the elements in the set of sensitivity indices, sort them, and calculate the cumulative contribution degree. Based on the parameter to be optimized corresponding to the sensitivity index when the cumulative contribution degree exceeds the threshold, construct a second parameter vector to be optimized.

[0104] An optimization parameter prediction module, configured to obtain all parameters and features and perform prediction through multi-objective modeling, feature extraction, and an XGBoost regression model to obtain a parameter value vector of the second parameter vector to be optimized;

[0105] Further, the steps of constructing the optimization parameter prediction model include:

[0106] Step S1: Establish an objective comprehensive evaluation mechanism to obtain multi-objective scoring labels. The objective comprehensive evaluation mechanism includes a heat exchange objective, an energy-saving objective, and an equipment status objective; set initial weights through the AHP analytic hierarchy process and use an adaptive fuzzy inference system to adjust the weight values in real time;

[0107] Step S2: Obtain historical environmental data, operating parameters, and equipment status data, and capture the equipment response characteristics in the time series through a multi-layer LSTM network and an attention enhancement mechanism to obtain equipment features;

[0108] Step S3: Train the XGBoost regression model;

[0109] Step S4: Obtain the trained XGBoost regression model and perform prediction. The model inputs include environmental features, equipment features, historical parameters, and multi-objective scoring labels; the model output includes the parameter value vector of the second parameter vector to be optimized.

[0110] A distributed control module, configured to obtain the parameter value vector of the second parameter vector to be optimized and perform optimization control of the parameters; the distributed parameter optimization control model includes setting local agents for the parameters to be optimized and performing parameter optimization control through a reinforcement learning algorithm.

[0111] Further, the steps of constructing the distributed parameter optimization control model include:

[0112] Step S10: Configure a local agent for each parameter and perform local control using a deep reinforcement learning algorithm; the state space of the local agent includes current environmental data, equipment operating status, and status information of the parameter value vector; the action space includes the adjustment range of the corresponding parameter;

[0113] Step S20: Construct a reward function for the local agent based on the heat exchange efficiency, energy consumption, and parameter change range;

[0114] Step S30: Train according to the deep reinforcement learning algorithm to obtain the trained local agent;

[0115] Further, reinforcement learning algorithms such as the PPO algorithm, DDPG algorithm, and TD3 algorithm.

[0116] Step S40: Obtain the parameter value vectors of the trained local agent and the second parameter vector to be optimized. According to the optimized parameter values and the current actual parameter values, if the deviation exceeds the set threshold, adopt a progressive adjustment strategy;

[0117] Step S50: Obtain the cumulative reward of the local agent. After reaching the set convergence condition of the objective function or reaching the maximum number of iterations, stop the iteration, and the control execution of the local agent ends.

[0118] The system proposed by the present invention realizes real-time intelligent optimization control of the operating parameters of the air condenser through the control frequency optimization module, the parameter to be optimized selection module, the optimized parameter prediction module, and the distributed control module, significantly enhancing the self-adaptability, robustness, and operating efficiency of the system.

[0119] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the control of operating parameters of an air condenser, characterized in that Including: Obtain and process environmental data, and construct an adaptive environmental monitoring value according to the thermal stress term, radiation term, and time-varying correction term; The calculation formula of the adaptive environmental monitoring value is: ； Among them, represents the adaptive environmental monitoring value, and represents the adaptive weight coefficient, represents the temperature, represents the temperature nonlinear coefficient, represents the sign function, represents the preset temperature, represents the humidity coupling coefficient, represents the humidity, represents the saturation humidity, represents the wind speed attenuation coefficient, represents the square root of the wind speed, represents the radiant heat intensity, and represents the pressure modulation parameter, represents the natural base, represents the atmospheric pressure, represents the standard atmospheric pressure, and represents the historical trend sensitivity coefficient, represents the hyperbolic tangent function, represents the temperature change trend; Construct a non-uniform frequency conversion strategy to optimize the operating parameters of the air condenser according to the transient response term, memory decay integral term, and non-linear compensation term of the adaptive environmental monitoring value; the calculation formula of the non-uniform frequency conversion strategy is: ； Among them, represents a non-uniform frequency conversion strategy, and represent the transient response term coefficient and the memory decay integral term coefficient, represents the derivative of the adaptive environmental monitoring value, represents Gaussian kernel convolution, represents the bandwidth adaptively adjusted according to the signal-to-noise ratio, represents the adaptive environmental monitoring value function, represents the memory decay coefficient, represents time, represents the integration variable, represents the non-linear compensation term coefficient, represents the sign function, represents the gradient operator, represents the Laplacian term; Obtain the structural characteristics and operating characteristics of the air condenser and obtain the first parameter vector to be optimized according to the principal component analysis method; obtain the first parameter vector to be optimized and perform sensitivity analysis, and obtain a sensitivity index set according to the sensitivity index; calculate the cumulative contribution degree according to the variance and expectation and sort, and obtain the second parameter vector to be optimized according to the threshold; Construct an optimization parameter prediction model according to the target scoring mechanism, where the target comprehensive evaluation mechanism includes a heat transfer target, an energy-saving target, and a device status target; the optimization parameter prediction model obtains all the parameters and characteristics of the air condenser and makes a prediction to obtain the parameter value vector of the second parameter vector to be optimized; The distributed parameter optimization control model includes local agents for setting parameters to be optimized. The state space of the local agent includes the current environmental data, device operating status, and state information of the parameter value vector; the action space includes the adjustment range of the corresponding parameter; construct a reward function for the local agent according to the heat transfer efficiency, energy consumption, and parameter change range; The local agent obtains the parameter value vector of the second parameter vector to be optimized and performs parameter optimization control through a reinforcement learning algorithm.

2. The operating parameter optimization control method of an air condenser according to claim 1, characterized in that It also includes obtaining the parameter values in the first parameter vector to be optimized and calculating the sensitivity index of the parameter values through sensitivity analysis to obtain a sensitivity index set; obtaining the elements in the sensitivity index set and sorting and calculating the cumulative contribution degree, and constructing a second parameter vector to be optimized according to the parameter to be optimized corresponding to the sensitivity index when the cumulative contribution degree exceeds the threshold.

3. The operating parameter optimization control method of an air condenser according to claim 1, characterized in that, The steps of constructing the optimization parameter prediction model include: Step S1: Establish a target comprehensive evaluation mechanism to obtain multi-target scoring labels. The target comprehensive evaluation mechanism includes a heat transfer target, an energy-saving target, and a device status target; set the initial weight through the AHP hierarchical analysis method, and use an adaptive fuzzy inference system to adjust the weight value in real time; Step S2: Obtain historical environmental data, operating parameters, and device status data, and capture the device response characteristics in the time series through a multi-layer LSTM network and an attention enhancement mechanism to obtain device characteristics; Step S3: Train the XGBoost regression model; Step S4: Obtain the trained XGBoost regression model and make a prediction. The model input includes environmental characteristics, device characteristics, historical parameters, and multi-target scoring labels; the model output includes the parameter value vector of the second parameter vector to be optimized.

4. The operating parameter optimization control method of an air condenser according to claim 1, characterized in that The steps of constructing the distributed parameter optimization control model include: Step S10: Configure a local agent for each parameter and perform local control using a deep reinforcement learning algorithm; the state space of the local agent includes the current environmental data, device operating status, and state information of the parameter value vector; the action space includes the adjustment range of the corresponding parameter; Step S20: Construct a reward function for the local agent based on the heat exchange efficiency, energy consumption, and parameter change range; Step S30: Train according to the deep reinforcement learning algorithm to obtain the trained local agent; Step S40: Obtain the parameter value vectors of the trained local agent and the second parameter vector to be optimized. According to the optimized parameter values and the current actual parameter values, if the deviation exceeds the set threshold, adopt a progressive adjustment strategy; Step S50: Obtain the cumulative reward of the local agent. After reaching the set convergence condition of the objective function or reaching the maximum number of iterations, stop the iteration, and the control execution of the local agent ends.

5. An operating parameter optimization control system for an air condenser, characterized in that, It includes: A control frequency optimization module, including an adaptive environmental monitoring value calculation unit and a non-uniform frequency conversion strategy control unit; Among them, the adaptive environmental monitoring value calculation unit obtains and processes environmental data, and constructs an adaptive environmental monitoring value according to the thermal stress term, radiation term, and time-varying correction term; the calculation formula of the adaptive environmental monitoring value is: ； Among them, represents the adaptive environmental monitoring value, and represents the adaptive weight coefficient, represents the temperature, represents the temperature nonlinear coefficient, represents the sign function, represents the preset temperature, represents the humidity coupling coefficient, represents the humidity, represents the saturation humidity, represents the wind speed attenuation coefficient, represents the square root of the wind speed, represents the radiant heat intensity, and represents the pressure modulation parameter, represents the natural base, represents the atmospheric pressure, represents the standard atmospheric pressure, and represents the historical trend sensitivity coefficient, represents the hyperbolic tangent function, represents the temperature change trend; The non-uniform frequency conversion strategy control unit constructs a non-uniform frequency conversion strategy to optimize the operating frequency of the air condenser according to the transient response term, memory decay integral term, and non-linear compensation term of the adaptive environmental monitoring value; the calculation formula of the non-uniform frequency conversion strategy is: ； Among them, represents a non-uniform frequency conversion strategy, and represent the transient response term coefficient and the memory decay integral term coefficient, represents the derivative of the adaptive environmental monitoring value, represents Gaussian kernel convolution, represents the bandwidth adaptively adjusted according to the signal-to-noise ratio, represents the adaptive environmental monitoring value function, represents the memory decay coefficient, represents time, represents the integration variable, represents the non-linear compensation term coefficient, represents the sign function, represents the gradient operator, represents the Laplacian term; A parameter to be optimized selection module, which is used to obtain the structural characteristics and operating characteristics of the air condenser and obtain the first parameter vector to be optimized according to the principal component analysis method; obtain the first parameter vector to be optimized and conduct a sensitivity analysis, and obtain a sensitivity index set according to the sensitivity index; calculate the cumulative contribution degree according to the variance and expectation and sort it, and obtain the second parameter vector to be optimized according to the threshold; An optimized parameter prediction module constructs an optimized parameter prediction model according to the target scoring mechanism, where the target comprehensive evaluation mechanism includes a heat exchange target, an energy saving target, and an equipment status target; the optimized parameter prediction model obtains all the parameters and characteristics of the air condenser and conducts a prediction to obtain the parameter value vector of the second parameter vector to be optimized; The distributed control module includes a distributed parameter optimization control model, which is used to set local agents for the parameters to be optimized. The state space of the local agent includes the current environmental data, equipment operating status, and state information of the parameter value vector; the action space includes the adjustment range of the corresponding parameters; construct a reward function for the local agent according to the heat exchange efficiency, energy consumption, and parameter change range; obtain the parameter value vector of the second parameter vector to be optimized and conduct parameter optimization control through the reinforcement learning algorithm.

Images