A building air conditioning system energy-saving operation simulation and optimization method and system

Abstract

The application provides a building air conditioning system energy-saving operation simulation and optimization method and system, and belongs to the technical field of building energy saving and system optimization. The method comprises the following steps: obtaining target building multi-source data, constructing and initializing a building air conditioning simulation model; obtaining meteorological prediction data and building use plan data of a next control period, inputting the building air conditioning simulation model to obtain predicted operation data; constructing a comprehensive objective function by minimizing total operation energy consumption and total operation cost, establishing an optimization control model by taking the indoor environment parameters in the comfort interval and the equipment operation parameters in the safety interval as constraints, and solving the optimization control model by using an optimization algorithm to obtain an optimal control instruction sequence of the equipment in the next control period and execute the optimal control instruction sequence, collecting actual operation data and calculating the deviation of the actual operation data from the predicted operation data to obtain an operation effect deviation value; and if the operation effect deviation value is greater than a deviation threshold value, calibrating the building air conditioning simulation model and / or iteratively updating the optimization control model and applying the optimization control model to a subsequent control period.

Description

Technical Field

[0001] This application relates to the field of building energy conservation and system optimization technology, and in particular to a method and system for energy-saving operation simulation and optimization of building air conditioning systems. Background Technology

[0002] Currently, there are two main methods for addressing energy-saving operation of building air conditioning systems. One method involves controlling the operation of air conditioning equipment based on fixed preset parameters, setting the on / off times and temperature ranges according to different times of day. The other method utilizes sensors to monitor indoor environmental parameters, automatically adjusting the operating status of the air conditioning equipment when these parameters exceed preset ranges. However, due to the inability to obtain multi-source data for the target building and construct an accurate building air conditioning simulation model, it is impossible to accurately predict the operating data for the next control cycle. Furthermore, the lack of a comprehensive objective function and corresponding optimized control model designed to minimize total operating energy consumption and total operating cost prevents the determination of the optimal control command sequence for the equipment. Moreover, the inability to calibrate and update the model based on the deviation between actual and predicted operating data makes it unable to adapt to changes in the building environment and usage conditions, thus failing to achieve highly efficient and energy-saving operation of the building air conditioning system.

[0003] Therefore, there is an urgent need for a simulation and optimization method and system for energy-saving operation of building air conditioning systems. Summary of the Invention

[0004] To address the aforementioned technical issues, this application provides a method and system for simulating and optimizing the energy-saving operation of a building air conditioning system.

[0005] A first aspect of this application provides a method for simulating and optimizing the energy-saving operation of a building air conditioning system, comprising: Acquire multi-source data of the target building, and construct and initialize a building air conditioning simulation model based on the multi-source data; Obtain meteorological forecast data and building usage plan data for the next control cycle, input them into the building air conditioning simulation model for forward simulation, and obtain the predicted operating data corresponding to the next control cycle; A comprehensive objective function is constructed to minimize total operating energy consumption and total operating cost. An optimal control model is established with indoor environmental parameters in the comfort range and equipment operating parameters in the safe range as constraints. The optimal control model is solved using an optimization algorithm to obtain the optimal control command sequence for the equipment in the next control cycle. Execute the optimal control command sequence, collect actual operating data, calculate the deviation between the actual operating data and the predicted operating data, and obtain the operating effect deviation value; If the deviation value of the operating effect is greater than the preset deviation threshold, then based on the historical actual operating data of multiple control cycles and the preset decision logic, it is decided to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model. The calibrated building air conditioning simulation model and / or the updated optimized control model are applied to subsequent control cycles.

[0006] A second aspect of this application provides a simulation and optimization system for energy-saving operation of a building air conditioning system, comprising: A model building module is used to acquire multi-source data of the target building, and to build and initialize a building air conditioning simulation model based on the multi-source data; The simulation prediction module is used to acquire meteorological forecast data and building usage plan data for the next control cycle, input them into the building air conditioning simulation model for forward simulation, and obtain the predicted operation data corresponding to the next control cycle. A model module is established to construct a comprehensive objective function that minimizes total operating energy consumption and total operating cost. An optimization control model is established with indoor environmental parameters in the comfort range and equipment operating parameters in the safe range as constraints. The optimization solution module is used to solve the optimized control model using optimization algorithms to obtain the optimal control command sequence for the equipment in the next control cycle; The operation deviation module is used to execute the optimal control command sequence, collect actual operation data, calculate the deviation between the actual operation data and the predicted operation data, and obtain the operation effect deviation value. The decision update module is used to determine, based on historical actual operating data of multiple control cycles and preset decision logic, to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model if the deviation value of the operating effect is greater than a preset deviation threshold. The model application module is used to apply the calibrated building air conditioning simulation model and / or the updated optimized control model to subsequent control cycles.

[0007] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described building air conditioning system energy-saving operation simulation and optimization method.

[0008] In a fourth aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for simulating and optimizing the energy-saving operation of a building air conditioning system.

[0009] The beneficial effects of the energy-saving operation simulation and optimization method and system for building air conditioning systems provided in this application are as follows: This application constructs and initializes a building air conditioning simulation model through multi-source data, performs forward simulation based on meteorological forecast data and building usage plan data, and can accurately predict the operating data of the next control cycle, solving the problem that the prior art cannot accurately predict operating data; it constructs a comprehensive objective function to minimize total operating energy consumption and cost and establishes an optimized control model, solves the optimal control command sequence of the equipment, and compares the actual and predicted operating data after executing the commands to obtain the operating effect deviation value. When the operating effect deviation value is greater than the preset deviation threshold, the simulation model is calibrated and / or the optimized control model is updated and applied to subsequent cycles, thereby realizing the simulation and optimization of energy-saving operation of building air conditioning systems, reducing total operating energy consumption and cost, and further improving indoor environmental comfort and equipment operation safety. Attached Figure Description

[0010] Figure 1 A flowchart illustrating a method for simulating and optimizing energy-saving operation of a building air conditioning system according to an embodiment of this application; Figure 2 A structural block diagram of a building air conditioning system energy-saving operation simulation and optimization system provided in an embodiment of this application; Figure 3 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0011] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0012] To make the purpose, technical solution, and advantages of this application clearer, the following will be described in conjunction with the appendix. Figure 1-3 The following is an explanation using specific examples.

[0013] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for simulating and optimizing energy-saving operation of a building air conditioning system according to an embodiment of this application. The method includes: S101: Obtain multi-source data of the target building, and construct and initialize the building air conditioning simulation model based on the multi-source data.

[0014] In this embodiment, multi-source data refers to various heterogeneous data sets used to construct the building air conditioning simulation model, involving dimensions such as building structure, equipment performance, historical operation, and meteorological environment. These include: building thermal parameters, air conditioning equipment performance parameters, historical operation data, meteorological data, and basic building information. Building thermal parameters include external wall / roof heat transfer coefficients, window shading coefficients, etc.; air conditioning equipment performance parameters include chiller COP, fan coil unit rated airflow, etc.; historical operation data includes hourly energy consumption, indoor temperature and humidity, equipment start-up and shutdown status, etc.; meteorological data includes historical outdoor temperature, humidity, solar radiation intensity, etc.; and basic building information includes building area, air-conditioned area, number of floors, etc.

[0015] Building air conditioning simulation models are digital simulation platforms built upon the principles of building heat balance and equipment performance characteristics. They can simulate changes in building heating and cooling loads and the energy consumption response of air conditioning systems under different meteorological conditions and usage scenarios, serving as tools for subsequent forward simulation and optimization control. Initialization involves inputting multi-source data into the building air conditioning simulation model, and through parameter assignment, model training, and error calibration, ensuring that the deviation between the initial predicted values ​​and historical actual values ​​of the building air conditioning simulation model is controlled within a preset range, such as less than or equal to 5%, thus enabling the building air conditioning simulation model to possess basic simulation capabilities.

[0016] S102: Obtain meteorological forecast data and building usage plan data for the next control cycle, input them into the building air conditioning simulation model for forward simulation, and obtain the predicted operation data corresponding to the next control cycle.

[0017] In this embodiment, the next control cycle is the minimum scheduling time unit for energy-saving optimization of the building's air conditioning system. It can be set to 1 day, 8 hours, or 1 hour according to actual operational needs. It serves as the time boundary between simulation prediction and command execution. In this embodiment, the default control cycle is 24 hours (00:00 to 24:00 the next day). Meteorological forecast data consists of hourly outdoor environmental forecast parameters within the next control cycle. These are external factors affecting the building's cooling and heating loads, including outdoor temperature, relative humidity, solar radiation intensity, wind speed, and other data, sourced from a meteorological forecast platform or the building's local meteorological forecast model. Building usage plan data is preset data representing the building's usage status within the next control cycle, including passenger flow forecast curves (generated based on patterns such as holidays and promotional activities), personnel work and rest times, equipment operation plans, etc., used to calculate the building's internal disturbance loads (personnel heat dissipation, equipment heat dissipation) and the optimal start-up time for pre-cooling / pre-heating before opening.

[0018] In this embodiment, forward simulation involves substituting input meteorological forecast data and building usage plan data into an initialized building air conditioning simulation model, and then forward-engineering the changes in building cooling and heating loads and the energy consumption response of the air conditioning system within the next control cycle according to the time series. The predicted operating data is the output of the forward simulation, including two types of curves: one is the hourly cooling / heating load prediction curve, representing the changes in cooling / heating demand at different times; the other is the corresponding energy consumption prediction curve, representing the hourly energy consumption of each device in the air conditioning system and the total energy consumption.

[0019] S103: Construct a comprehensive objective function by minimizing total operating energy consumption and total operating cost, and establish an optimization control model with indoor environmental parameters in the comfort range and equipment operating parameters in the safe range as constraints.

[0020] In this embodiment, the comprehensive objective function is a mathematical function constructed with minimizing total operating energy consumption and minimizing total operating cost as dual optimization objectives. Weighting coefficients balance the priorities of the two objectives, adapting to different operational strategies, such as energy efficiency priority, cost priority, and energy efficiency-cost balance, serving as a guide for the optimization control model. Total operating energy consumption is the sum of energy consumption of all equipment in the air conditioning system (chillers, pumps, cooling towers, fan coil units, etc.) in the next control cycle, expressed in kWh, representing the system's energy consumption level. Total operating cost is the total electricity cost of the air conditioning system in the next control cycle, calculated based on the time-of-use pricing policy, expressed in yuan, and its value is the sum of the products of energy consumption and corresponding electricity price for each time period.

[0021] In this embodiment, the comfortable range of indoor environmental parameters refers to the range of indoor physical parameters that meet human comfort, including indoor temperature (24-28℃) and relative humidity (40%-60%), which are hard constraints to ensure user experience. The safe range of equipment operating parameters refers to the parameter threshold range for stable and safe operation of air conditioning equipment, such as chilled water outlet temperature of chiller units (7-12℃), water pump operating frequency (25-50Hz), and equipment start-up and shutdown frequency (less than or equal to 6 times / day). Exceeding these ranges can lead to equipment failure or reduced lifespan. The optimized control model is a mathematical optimization model composed of a comprehensive objective function and constraints. Solving it through an optimization algorithm yields a sequence of equipment control commands that satisfies the constraints and optimizes the objective function.

[0022] S104: Solve the optimal control model using an optimization algorithm to obtain the optimal control command sequence for the equipment in the next control cycle.

[0023] In this embodiment, the optimization algorithm is a numerical calculation method used to solve the optimization control model. The goal is to find the optimal combination of decision variables that maximizes the comprehensive objective function (minimizing energy consumption and cost) while satisfying all constraints. Based on the continuous and multi-constraint characteristics of building air conditioning system optimization, this embodiment employs the particle swarm optimization algorithm (which has fast convergence speed and is suitable for multi-variable optimization scenarios).

[0024] Specifically, the Particle Swarm Optimization (PSO) algorithm is used to solve the optimal control model to output the optimal control command sequence for the equipment. It adopts a three-level hierarchical structure: algorithm control layer, particle search layer, and command output layer. The layers form a closed loop through data feedback: The algorithm control layer, as the top-level unit, is responsible for configuring the core parameters of the algorithm, setting the iteration termination conditions, and coordinating the global search direction of the population; The particle search layer is the core execution unit. Each particle corresponds to a complete set of equipment control command sequences, such as chiller load rate, water pump operating frequency, and air conditioning terminal air volume. Particles adjust the values ​​of command parameters by tracking their own historical best command sequences and the global best command sequences of the population; The command output layer is the bottom-level execution unit. It is responsible for substituting the control command sequence corresponding to the particle into the optimal control model, calculating the objective function value (such as total operating cost and total energy consumption), and feeding it back to the particle search layer to provide a basis for particle position updates.

[0025] In this embodiment, the optimized control model is a mathematical model constructed including a comprehensive objective function and constraints. It serves as the basis for optimization solutions. The inputs are decision variables (equipment control parameters), and the output is the objective function value (a weighted sum of energy consumption and cost). The solution process must strictly meet the constraints of indoor comfort and equipment safety. The optimal equipment control command sequence is a set of equipment control parameters obtained after the optimization algorithm solves the problem, covering all time steps of the next control cycle. It is a standardized command to drive the air conditioning system. It includes chiller load rate, water pump operating frequency, cooling tower fan start / stop status, fan coil unit fan speed at each moment, etc., arranged in a time sequence to form a command sequence that can be directly imported into the air conditioning control system for execution.

[0026] S105: Execute the optimal control command sequence, collect actual operating data, calculate the deviation between the actual operating data and the predicted operating data, and obtain the operating effect deviation value.

[0027] In this embodiment, actual operating data refers to the real operating parameters collected by sensors and the Building Energy Management System (BEMS) after the air conditioning system executes the optimal control command. These include the actual hourly cooling / heating load curves of the building, the actual hourly energy consumption curves of each device in the air conditioning system, and the actual indoor temperature and humidity values. Predicted operating data is the prediction result obtained through forward simulation of the building air conditioning simulation model, corresponding one-to-one with the actual operating data. This includes hourly cooling / heating load prediction curves and energy consumption prediction curves, serving as a reference for deviation calculation. The operating performance deviation value is an evaluation index that comprehensively quantifies energy consumption prediction deviation, economic prediction deviation, and temperature prediction deviation. It is obtained by weighted summation of multi-dimensional deviation indicators and is a key threshold for determining whether to calibrate the simulation model or iteratively optimize the control model.

[0028] S106: If the deviation value of the operating effect is greater than the preset deviation threshold, then based on the historical actual operating data of multiple control cycles and the preset decision logic, it is decided to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model.

[0029] In this embodiment, the preset deviation threshold is a pre-defined critical value for operational performance deviation. It serves as the standard for determining whether model calibration or iteration is necessary. The value is determined based on the target building's operational accuracy requirements and historical optimization results, and is set at 3%-5%. A value exceeding this threshold indicates that the prediction / optimization accuracy of the building air conditioning simulation model or optimized control model can no longer meet actual needs. Historical actual operational data from multiple control cycles refers to the real operational data of the air conditioning system collected over several consecutive control cycles (e.g., the last 7 days, the last 30 days), including hourly cooling / heating load, energy consumption, indoor temperature and humidity, and equipment operating parameters. This data forms the basis for analyzing the root causes of deviations and supporting model calibration / iteration. The preset decision logic is a pre-defined model adjustment rule based on the contribution analysis of deviation indicators. It determines the root cause of deviations—whether it's a simulation model malfunction or unreasonable optimized control model parameters—by judging the proportion and trend of energy consumption deviation, economic deviation, and temperature deviation. This leads to a decision to calibrate the simulation model, iterate the optimized control model, or adjust both simultaneously.

[0030] In this embodiment, the calibration of the building air conditioning simulation model addresses the problem of excessive deviation between the predicted and actual values ​​of the simulation model. It involves correcting parameters such as building thermal parameters and equipment performance curves through inversion, thereby improving the accuracy of building air conditioning simulation predictions. This is suitable for scenarios where energy consumption and temperature deviations are dominant. The iterative update of the optimized control model addresses the problems of mismatch between the optimization objective and actual needs, and unreasonable constraint boundaries. It involves adjusting the weight coefficients of the comprehensive objective function and the boundary parameters of the constraint parameters. This is suitable for scenarios where economic deviations are dominant.

[0031] S107: Apply the calibrated building air conditioning simulation model and / or the updated optimized control model to subsequent control cycles.

[0032] In this embodiment, the calibrated building air conditioning simulation model addresses the problem of excessive prediction deviation in the original simulation model. By correcting parameters such as building thermal parameters and equipment performance curves, it achieves an iterative version with improved accuracy. The deviation between its predicted values ​​and actual operating data is now controlled within a preset range, providing more accurate load and energy consumption prediction capabilities. The updated optimized control model addresses the issues of unreasonable objective function weights and inaccurate constraint boundaries in the original optimization model. It is a new version formed by adjusting the dual-objective weight coefficients or constraint parameter boundaries, better aligning with actual building operation needs, such as prioritizing energy efficiency and cost, and outputting superior equipment control commands. Subsequent control cycles are the next one or more control cycles after triggering the model calibration / update operation. The default cycle is 24 hours, serving as the time unit for verifying the closed-loop optimization of the air conditioning system after iteration.

[0033] As can be seen from the above, this application constructs and initializes a building air conditioning simulation model through multi-source data, and performs forward simulation based on meteorological forecast data and building usage plan data, which can accurately predict the operating data of the next control cycle, solving the problem that existing technologies cannot accurately predict operating data. By constructing a comprehensive objective function to minimize total operating energy consumption and cost and establishing an optimized control model, the optimal control command sequence of the equipment is obtained by solving the problem. After executing the command, the actual and predicted operating data are compared to obtain the operating effect deviation value. When the operating effect deviation value is greater than the preset deviation threshold, the simulation model is calibrated and / or the optimized control model is updated and applied to subsequent cycles. This realizes the simulation and optimization of energy-saving operation of the building air conditioning system, reduces total operating energy consumption and cost, and makes the indoor environment comfortable and the equipment safe to operate.

[0034] In one embodiment of this application, the deviation between actual operating data and predicted operating data is calculated to obtain an operating performance deviation value, including: The relative error between the actual total energy consumption in the actual energy consumption curve and the corresponding predicted total energy consumption in the energy consumption prediction curve is calculated and used as an energy consumption deviation index. The actual total electricity cost is calculated based on time-of-use electricity price data and actual energy consumption curves. The predicted total electricity cost is calculated based on the energy consumption prediction curve and time-of-use electricity price data. The relative error between the actual total electricity cost and the predicted total electricity cost is calculated as an indicator of economic deviation. The average absolute error between the average actual indoor temperature and the predicted indoor temperature of the target building is calculated and used as a temperature deviation index. The weighted sum of the energy consumption deviation index, economic deviation index, and temperature deviation index is used as the operational performance deviation value.

[0035] In this embodiment, the energy consumption deviation index represents the relative error value between the actual total energy consumption and the predicted total energy consumption of the air conditioning system. It indicates the accuracy of the building air conditioning simulation model's energy consumption prediction; the smaller the value, the higher the prediction accuracy. Time-of-use (TOU) electricity price data is a differentiated electricity price standard established by the power sector based on the differences in electricity load at different times. It is divided into three periods: peak price, flat price, and off-peak price. The peak period price is the highest, and the off-peak period price is the lowest, serving as the basis for calculating the operating electricity cost of the air conditioning system. The actual total electricity cost is the real electricity expenditure calculated based on the actual energy consumption curve and TOU electricity price data after the air conditioning system executes the optimal control command. The predicted total electricity cost is the electricity expenditure pre-calculated based on the energy consumption prediction curve output by the building air conditioning simulation model and the TOU electricity price data, serving as a reference benchmark for evaluating economic deviation.

[0036] In this embodiment, the economic deviation index is the relative error between the actual total electricity cost and the predicted total electricity cost, representing the accuracy of the optimization control model in cost prediction and optimization. The temperature deviation index is the average absolute error between the average actual indoor temperature and the predicted indoor temperature of the target building, representing the accuracy of the simulation model's prediction of the indoor thermal environment, and directly related to the comfort experience of people inside the building. The operational performance deviation value is a comprehensive evaluation value obtained by weighted summation of the energy consumption deviation index, economic deviation index, and temperature deviation index, and serves as the basis for determining whether the building air conditioning simulation model and optimization control model need calibration / iteration.

[0037] As can be seen from the above, this embodiment calculates the energy consumption deviation index, economic deviation index, and temperature deviation index, and then weights and sums the three to obtain the operational performance deviation value. This can comprehensively consider the energy consumption, economic cost, and indoor temperature of the building air conditioning system, providing an accurate basis for determining whether the model needs to be calibrated and updated.

[0038] In one embodiment of this application, if the operational performance deviation value is greater than a preset deviation threshold, then based on historical actual operational data from multiple control cycles and preset decision logic, it is decided to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model, including: Obtain the operational performance deviation values ​​for the current and historical control cycles, along with their corresponding energy consumption deviation, economic performance deviation, and temperature deviation indicators. The analysis includes the number of cycles in which the operational performance deviation value exceeds the preset deviation threshold, the slope of the trend, and the contribution of each of the energy consumption deviation index, economic deviation index, and temperature deviation index to the total deviation. If the average contribution ratio of the energy consumption deviation index is greater than the contribution threshold for three consecutive control cycles and the slope of its trend is greater than the preset slope threshold, then the building air conditioning simulation model is calibrated. If the average contribution ratio of the economic deviation index is greater than the contribution threshold for three consecutive control cycles, the weight coefficients of the comprehensive objective function and the boundary parameters of the constraint conditions in the optimized control model will be adjusted. If the temperature deviation index exceeds 1.5 times its preset deviation threshold for two consecutive control cycles, the room temperature prediction module of the building air conditioning simulation model will be calibrated and the comfort range constraints in the optimized control model will be reviewed and updated.

[0039] In this embodiment, the contribution threshold is a critical percentage used to determine whether a certain type of deviation indicator is the dominant factor in the total deviation. The determination of the contribution threshold is based on the statistical patterns of historical operating data and the needs of engineering practice. The steps are as follows: First, calculate the contribution of the energy consumption deviation indicator. The formula for calculating the contribution of the energy consumption deviation indicator to the operational performance deviation value is: C = (ω*N / D)*100%, where C is the contribution of the energy consumption deviation indicator, ω is the energy consumption deviation weight coefficient, N is the energy consumption deviation indicator value, and D is the operational performance deviation value. Second, statistically analyze the historical contribution benchmark value. Retrieve historical operating data from the target building for the past 60-90 control cycles and calculate the contribution C of the energy consumption deviation indicator for each cycle to obtain the historical contribution sequence. Finally, calculate the contribution threshold. Perform statistical analysis on the historical contribution sequence and calculate the 75th quartile Q of the sequence, i.e., the upper quartile. This is used to divide the data distribution interval, indicating that 75% of the data in the dataset is less than or equal to this value, and 25% of the data is greater than or equal to this value. Adjustments are made based on engineering experience: if energy efficiency is the core operational objective of a building, the contribution threshold is increased by 10%-15%; if economic efficiency is the core objective, the contribution threshold is decreased by 5%-10%. The final contribution threshold ranges from 30% to 50%. For example, if Q is statistically obtained as 35%, after adjustment based on the energy efficiency priority strategy, the contribution threshold is set at 40%.

[0040] In this embodiment, the slope of the trend is the slope of a linear fit calculated with the control period as the horizontal axis and the deviation index value as the vertical axis. It represents the trend of the deviation index; a positive slope indicates a continuous increase in deviation, while a negative slope indicates a continuous decrease in deviation. This serves as the basis for judging whether the inaccuracy of the building air conditioning simulation model has worsened. Building air conditioning simulation model calibration is a process that reduces the deviation between predicted and actual values ​​by correcting elements in the model, such as building thermal parameters, equipment performance curves, and room temperature prediction modules, specifically for scenarios dominated by energy consumption or temperature deviations.

[0041] In this embodiment, adjusting the weight coefficients of the comprehensive objective function addresses scenarios dominated by economic deviations. It changes the weight ratio between minimizing energy consumption and minimizing cost, making the optimization objective more aligned with actual operational needs. For example, prioritizing cost increases the weight of the cost objective. Adjusting the constraint boundary parameters also addresses scenarios dominated by economic deviations. It corrects the constraint ranges of indoor environmental parameters and equipment operating parameters in the optimization control model. For example, adjusting the chilled water temperature boundary of the chiller unit eliminates cost deviations caused by overly strict / loose constraints. The room temperature prediction module calibration is a sub-module calibration in the building air conditioning simulation model used to optimize the accuracy of indoor temperature prediction. It improves temperature prediction accuracy by importing historical actual room temperature data to correct the prediction algorithm.

[0042] As can be seen from the above, this embodiment makes targeted adjustments to the building air conditioning simulation model and the optimized control model based on different indicators. Acquiring and analyzing the data for each indicator clarifies the source and extent of deviations. Calibrating the building air conditioning simulation model when the energy consumption deviation indicator meets the conditions can correct errors in building thermal parameters or equipment performance models. Adjusting the weighting coefficients and boundary parameters of the optimized control model when the economic deviation indicator meets the conditions makes the model more consistent with actual business needs. Calibrating the room temperature prediction module and reviewing and updating the comfort range constraints when the temperature deviation indicator meets the conditions ensures indoor temperature comfort and model accuracy, thereby improving the effectiveness of energy-saving simulation and optimization of the building air conditioning system.

[0043] In one embodiment of this application, calibrating a building air conditioning simulation model includes: Compare and analyze the deviation between the predicted energy consumption curve and the actual energy consumption curve to identify the error sources that cause the deviation; error sources include inaccurate building thermal parameters or inaccurate equipment performance models. Based on the error sources, construct a calibration dataset from historical operational data and perform calibration operations: If the source of error is inaccurate building thermal parameters, then an optimization algorithm is used to invert and update the parameters of the building air conditioning simulation model. If the error source is an inaccurate equipment performance model, then the historical actual operating data in the calibration dataset is used as the training target, and the neural network is used to correct the equipment performance mapping relationship in the original building air conditioning simulation model.

[0044] In this embodiment, the error source is the fundamental reason for the deviation between the predicted and actual values ​​of the building air conditioning simulation model. This embodiment identifies two types of factors: inaccurate building thermal parameters (e.g., the heat transfer coefficient of external walls and the shading coefficient of windows do not match reality) and inaccurate equipment performance models (e.g., the COP curve of chillers and the power curve of water pumps do not match the actual operating characteristics of the equipment). The calibration dataset is a subset of effective data selected from historical actual operating data, covering different meteorological conditions and building usage scenarios. It includes hourly meteorological data, building load data, equipment operating parameters, and actual energy consumption values, serving as the training and verification basis for the building air conditioning simulation model calibration. The optimization algorithm inversion update aims to minimize the deviation between the energy consumption output of the building air conditioning simulation model and the actual energy consumption. This embodiment uses a genetic algorithm for inversion update, calculating the optimal correction values ​​of building thermal parameters through reverse derivation, thereby reducing the model prediction deviation.

[0045] Specifically, the genetic algorithm updates the building thermal parameters by retrieving the error sources. It adopts a three-level hierarchical structure: a population management layer, a chromosome coding layer, and a fitness calculation layer. The layers have a closed-loop logic of top-down instruction transmission and bottom-up result feedback. The population management layer, as the top-level unit of the algorithm, is responsible for initializing the population size, setting iterative evolution rules, executing core evolutionary operations such as selection, crossover, and mutation, and coordinating the global search direction of the algorithm. The chromosome coding layer is the core execution unit of the algorithm. It uses real number encoding. Each chromosome corresponds to a set of building thermal parameters to be inverted, such as the heat transfer coefficient of the exterior wall, the heat transfer coefficient of the roof, and the shading coefficient of the windows. Each gene locus corresponds to a specific value of the thermal parameter. The fitness calculation layer is the bottom-level data unit of the algorithm. It is responsible for substituting the thermal parameters corresponding to the chromosomes into the building air conditioning simulation model, calculating the deviation between the simulated energy consumption and the actual energy consumption as the fitness, and feeding it back to the population management layer to guide the evolution of the next generation of the population, so that the inverted parameters fit the actual building energy consumption characteristics.

[0046] In this embodiment, the equipment performance mapping relationship describes the mathematical correspondence between the operating parameters and performance indicators of air conditioning equipment, and serves as the basis for the building air conditioning simulation model to calculate equipment energy consumption. Neural network correction utilizes the nonlinear fitting capability of neural networks, using a calibration dataset as training samples, to learn the true performance mapping relationship of the equipment, replacing the original inaccurate mathematical model and improving the accuracy of equipment energy consumption prediction.

[0047] Specifically, the neural network used to correct the performance mapping relationship of the building air conditioning simulation model adopts a three-level fully connected hierarchical structure of input layer-hidden layer-output layer. The layers achieve linear transformation and nonlinear mapping of data through weight matrices. The functions and relationships of each layer are clearly defined: The input layer is the information receiving unit of the neural network, with the number of neurons matching the dimensions of the influencing factors of equipment performance. Specifically, it includes three neurons: chiller load rate, chilled water outlet temperature, and cooling water inlet temperature, which are responsible for receiving the equipment operating parameters in the calibration dataset; The hidden layer is the core feature extraction unit of the neural network, with one hidden layer and 12 neurons. The activation function is the Sigmoid function, which is used to explore the nonlinear correlation between input parameters and equipment performance indicators; The output layer is the result output unit of the neural network, with one neuron, which is responsible for outputting the real-time energy efficiency value (COP) of the equipment. The three-layer structure forms a mapping link from top to bottom of parameter input-feature extraction-performance output, accurately fitting the actual performance mapping relationship of the equipment.

[0048] As can be seen from the above, this embodiment identifies the error source by comparing and analyzing the deviation between the energy consumption prediction curve and the actual curve. It uses optimization algorithms to update the parameters of the building air conditioning simulation model and corrects the mapping relationship by training a neural network with historical actual operating data to correct the inaccuracy of building thermal parameters. This effectively calibrates the building air conditioning simulation model, improves the accuracy of the building air conditioning simulation model, and provides a more reliable foundation for subsequent energy-saving operation simulation and optimization of the air conditioning system.

[0049] In one embodiment of this application, adjusting the weight coefficients of the integrated objective function and the boundary parameters of the constraints in the optimization control model includes: Multidimensional regression analysis was performed on the data of economic deviation indicators to identify the sources of deviation. If the source of deviation is that the boundary parameter settings of the constraint conditions do not match the actual operating range, then the boundary parameters in the constraint conditions should be corrected. If the source of deviation is that the weight coefficients of the two terms in the comprehensive objective function, namely minimizing total operating energy consumption and minimizing total operating cost, do not match the priority of actual business needs, then the weight coefficients of the corresponding objective terms in the comprehensive objective function should be adjusted.

[0050] In this embodiment, multidimensional regression analysis is a statistical analysis method that uses the economic deviation index as the dependent variable and the key parameters of the optimization control model as independent variables, such as the objective function weight coefficients, constraint boundary values, and time-of-use electricity price fluctuation coefficients, to construct a multiple regression equation. This method is used to represent the degree of influence of each parameter on the economic deviation, thereby locating the root cause of the deviation. The deviation source is the reason why the economic deviation index exceeds the acceptable range. In this embodiment, there are two types: First, the constraint boundary parameters do not match the actual operating range, such as overly strict / loose constraints on equipment operating parameters, causing the optimized solution to deviate from the actual optimum; second, the objective function weight coefficients do not match the priority of business needs, such as excessively high energy consumption weights under a cost-first strategy, resulting in insufficient cost optimization.

[0051] In this embodiment, the constraint boundary parameters are the threshold ranges that limit the operating parameters of the equipment and the indoor environment in the optimization control model. Examples include the upper and lower limits of the chilled water outlet temperature of the chiller unit, the operating frequency range of the water pump, and the comfortable indoor temperature range. These are hard constraints for the optimization solution. The weight coefficients of the comprehensive objective function are proportional coefficients that balance the two objectives of minimizing total operating energy consumption and minimizing total operating cost. The magnitude of the weight value directly indicates the priority of the operating strategy; cost priority results in a higher cost weight, and energy efficiency priority results in a higher energy consumption weight.

[0052] As can be seen from the above, this embodiment obtains the source of deviation by performing multidimensional regression analysis on the economic deviation index data, thus clarifying the cause of the deviation. If the source of deviation is that the boundary parameters of the constraint conditions do not match the actual operating range, the boundary parameters in the constraint conditions are corrected to make the constraint conditions of the optimization control model more in line with the actual operating conditions. If the source of deviation is that the weight coefficients of the two terms of minimizing total operating energy consumption and minimizing total operating cost in the comprehensive objective function do not match the priority of actual business needs, the weight coefficients of the corresponding objective terms in the comprehensive objective function are adjusted to make the target settings of the optimization control model more in line with actual business needs, thereby improving the effect of energy-saving operation simulation and optimization of building air conditioning systems and reducing operating energy consumption and costs.

[0053] In one embodiment of this application, if the source of deviation is that the boundary parameter settings of the constraint conditions do not match the actual operating range, then correcting the boundary parameters in the constraint conditions includes: Based on multidimensional regression analysis, constraints that are related to economic deviation indicators and have become active or tight constraints in historical optimization solutions are identified as constraints to be modified. Extract the data sequence of operating parameters corresponding to the constraints to be corrected from historical actual operating data; Based on the statistical distribution of the data sequence, a preset confidence quantile interval is calculated as a candidate boundary value; The intersection of the candidate boundary values ​​and the allowable parameter range of the building air conditioning system design code is used as the safety boundary parameter after the constraint conditions are updated.

[0054] In this embodiment, active or tight constraints are constraints where the optimal solution of the decision variables falls exactly on the constraint boundary during the solution process of the optimization control model. These constraints directly limit the further optimization space of the optimal solution and are the key constraint type that causes economic deviation. This is different from relaxed constraints, where the optimal solution does not touch the constraint boundary and has no impact on the optimization result. The constraints to be corrected are those selected through multidimensional regression analysis, which are related to the economic deviation index and have repeatedly become active or tight constraints in historical solutions. These are the objects of this boundary parameter correction.

[0055] In this embodiment, the confidence quantile interval is a range of parameter values ​​calculated based on the statistical distribution of historical operating parameter data sequences and a preset confidence level (e.g., 95%). It represents the actual reasonable operating range of the parameter. Commonly used quantiles are the 25th quantile (lower quartile) and 75th quantile (upper quartile), or the 5th quantile and 95th quantile. The allowable parameter range in the building air conditioning system design code is the safe and compliant range of equipment operating parameters and environmental parameters specified in national or industry standards (e.g., the "Energy Conservation Design Standard for Public Buildings"). It serves as the hard bottom line for constraining parameter correction, ensuring that the corrected parameters meet engineering safety requirements. The intersection range is the overlapping portion of the confidence quantile interval and the allowable range in the design code. It represents the final value range of the constraint boundary parameters, balancing actual operational rationality and safety compliance.

[0056] As can be seen from the above, this embodiment identifies the constraints to be corrected based on multidimensional regression analysis, extracts the data sequence of corresponding operating parameters from historical actual operating data, calculates candidate boundary values ​​and takes the intersection with the parameter range allowed by the design specifications to correct the boundary parameters of the constraints, so that the boundary parameters of the constraints are more in line with the actual operating range, improves the accuracy of the optimized control model, and enhances the economy and reliability of the energy-saving operation of the building air conditioning system.

[0057] In one embodiment of this application, if the source of deviation is that the weight coefficients of the two terms minimizing total operating energy consumption and minimizing total operating cost in the comprehensive objective function do not match the priority of actual business needs, then the weight coefficients of the corresponding objective terms in the comprehensive objective function are adjusted, including: Obtain the priority factor of actual business needs; the priority is determined by querying the preset operation strategy mapping table. The deviation rates of performance indicators for total operating energy consumption and total operating cost are calculated based on actual operating data. The deviation rates include the deviation rates of energy efficiency indicators and operating cost control indicators. If the deviation rate of the energy consumption efficiency index is greater than the deviation rate of the operating cost control index, and the priority factor indicates that energy efficiency takes precedence, then the weight of the target of minimizing total energy consumption is increased based on the first adjustment step size, and the weight of the target of minimizing total system operating cost is decreased based on the first adjustment step size. If the deviation rate of the operating cost control indicator is greater than the deviation rate of the energy consumption efficiency indicator, and the priority factor indicates that economic efficiency takes precedence, then the weight of the target of minimizing the total operating cost of the system is increased based on the second adjustment step size, and the weight of the target of minimizing the total energy consumption is decreased based on the second adjustment step size.

[0058] In this embodiment, the priority factor is an identifier parameter used to quantify the priority of actual business needs. It is a preset discrete value, such as 0 representing economic priority, 1 representing energy efficiency priority, and 2 representing balanced priority. It serves as the basis for adjusting the weighting coefficient. The operation strategy mapping table is a pre-defined table that associates different operational stages of a building with priority factors, including dimensions such as season, electricity consumption period, and operational goals. The priority corresponding to the current business can be directly queried. The performance indicator deviation rate is the relative deviation between the actual performance indicator and the target performance indicator. In this embodiment, it is divided into two categories: energy efficiency indicator deviation rate and operating cost control indicator deviation rate. Among them, the energy efficiency indicator deviation rate is the relative error between the actual energy efficiency and the target energy efficiency, indicating the degree of achievement of the energy consumption optimization goal; the operating cost control indicator deviation rate is the relative error between the actual operating cost and the target operating cost, indicating the degree of achievement of the cost optimization goal.

[0059] In this embodiment, the adjustment step size is the magnitude of a single adjustment of the weight coefficient, serving as a parameter to ensure the stability of the weight adjustment. The first adjustment step size is used for energy efficiency-first scenarios, and the second adjustment step size is used for economy-first scenarios. In the weight coefficient adjustment of the building air conditioning optimization control model, the core calculation of the first adjustment step size (energy efficiency-first scenario) and the second adjustment step size (economy-first scenario) is to match the model's stability and its sensitivity to business needs. This needs to be determined based on three dimensions: historical adjustment effects, deviation rate fluctuation range, and the rigidity of business priorities.

[0060] Specifically, the calculation of the first and second step lengths uses the coefficient of variation of historical deviation rates as the core factor, with the formulas: First step length = K1 * V1, Second step length = K2 * V2, where K1 and K2 are priority correction coefficients, rigidly set according to business priorities. For energy efficiency priority, K1 ranges from 0.05 to 0.1; for economic priority, K2 ranges from 0.08 to 0.12. V1 and V2 are the coefficients of variation of historical deviation rates, representing the dispersion of the deviation rate, with the formulas: V1 = σ / μ, V2 = σ / μ. Here, σ is the standard deviation of the deviation rate, and μ is the mean of the deviation rate. V1 calculates the historical sequence of deviation rates for energy efficiency indicators, and V2 calculates the historical sequence of deviation rates for operating cost control indicators. The constraint range for the first step length is 0.05-0.15, and the constraint range for the second step length is 0.08-0.2. If the calculated step size is not within the constraint range, a truncation correction is required. When it is greater than the upper limit, the step size is taken from the upper limit; when it is less than the lower limit, the step size is taken from the lower limit.

[0061] For example, to calculate the first adjustment step size (energy efficiency priority), select the energy efficiency index deviation rate sequence of the last 10 control cycles [3.2%, 4.5%, 3.8%, 5.1%, 4.0%, 3.5%, 4.8%, 3.0%, 4.2%, 3.6%], and calculate the coefficient of variation V1: the mean deviation rate μ = 3.2% + 4.5% + ... + 3.6% / 10 = 3.97%, the standard deviation of the deviation rate σ ≈ 0.65%, then V1 = 0.65 / 3.97 ≈ 0.164. Set the priority correction coefficient: energy efficiency priority, take K1 = 0.08. The first step size = 0.08 * 0.164 ≈ 0.0131, because it is less than the lower limit of constraint 0.05, so the first step size is corrected to 0.05.

[0062] As can be seen from the above, this embodiment obtains the priority factor of actual business needs, calculates the deviation rate of performance indicators for total operating energy consumption and total operating cost, and adjusts the weight coefficient of the corresponding objective item in the comprehensive objective function according to the relationship between the deviation rate of energy efficiency indicator and the deviation rate of operating cost control indicator, as well as the priority factor indication. This makes the operation of the building air conditioning system more in line with actual business needs, effectively balances the system's energy consumption and cost, and achieves optimization of energy saving and economical operation.

[0063] In one embodiment of this application, when calculating the energy consumption deviation index and the economic deviation index, a weighted calculation is performed using time-period error tolerance: The control period is divided into multiple time periods, including peak electricity price periods, flat electricity price periods, off-peak electricity price periods, and periods of significant changes in building personnel density; Different deviation tolerance coefficients are set for each time period, with the lowest tolerance coefficients for peak electricity price periods and economically sensitive periods; The relative errors of energy consumption and electricity cost for each time period are calculated separately, then multiplied by the tolerance coefficient of the corresponding time period for weighting, and finally summed to obtain the overall energy consumption deviation index and economic deviation index.

[0064] In this embodiment, the time-segmented error tolerance weighted calculation is based on the operational characteristics of different time periods (electricity price level, personnel density). Differentiated error tolerance coefficients are set for each time period, and the deviation values ​​calculated for each time period are weighted and summed to obtain the overall deviation index. This method ensures that deviations in highly sensitive time periods have a greater impact on the overall index. Dividing the control cycle into time periods involves splitting a complete control cycle (e.g., 24 hours) into sub-time periods with significant operational characteristics. In this embodiment, there are four categories: peak electricity price periods (highest electricity price), flat electricity price periods (medium electricity price), off-peak electricity price periods (lowest electricity price), and periods with significant changes in building personnel density, such as one hour before a shopping mall opens and one hour after it closes, where high personnel flow leads to large load fluctuations. The deviation tolerance coefficient is a weighted coefficient representing the acceptable degree of deviation in different time periods. The smaller the coefficient value, the more sensitive and intolerable the deviation is in that time period, and the higher its contribution to the overall index during weighted calculation; conversely, the larger the coefficient, the higher the deviation tolerance. Economically sensitive periods are those that react strongly to changes in operating costs and overlap with peak electricity price periods. During these periods, a 1% increase in energy consumption will lead to an increase in electricity prices, so a minimum deviation tolerance factor needs to be set.

[0065] As can be seen from the above, this embodiment divides the control cycle into peak electricity price periods, flat electricity price periods, off-peak electricity price periods, and periods with significant changes in building personnel density. It sets different deviation tolerance coefficients for different periods and calculates energy consumption deviation indicators and economic deviation indicators by weighting them. This can more accurately represent the impact of energy consumption and electricity cost errors on the overall indicators in different periods, especially highlighting the deviation impact of peak electricity price periods and economically sensitive periods, thereby improving the accuracy of operational performance evaluation.

[0066] In one embodiment of this application, the weighted sum of the energy consumption deviation index, the economic deviation index, and the temperature deviation index is used as the operational performance deviation value, including: Based on the target building's attributes and operating season, an initial set of weight coefficients is obtained by matching from a preset weight configuration database. The initial set of weight coefficients includes energy consumption deviation weight, economic deviation weight, and temperature deviation weight. The historical actual operation data of multiple control cycles, building usage plan data, and corresponding historical operation effect deviation values ​​are input into the weight adjustment model based on reinforcement learning to obtain the adjustment amount of the initial weight coefficient set. Based on the initial set of weight coefficients and the adjustment amount, the final weight coefficients are calculated. The energy consumption deviation index, economic deviation index, and temperature deviation index are weighted and summed using weighting coefficients to calculate the operational performance deviation value.

[0067] In this embodiment, the attributes of the target building are its classification features, including building type (commercial complex / office building / hospital / residential), functional positioning (profit-making / public welfare), and scale (building area, peak air conditioning load), which serve as the basis for matching the initial weights. The preset weight configuration database is a pre-built dataset storing weight coefficients corresponding to different building attribute-operating season combinations. It includes a large amount of historical project weight configuration experience, enabling direct matching and output of initial weights, ensuring the rationality of weight values. The reinforcement learning weight adjustment model is an adaptive optimization model built based on the reinforcement learning framework. This embodiment uses a deep Q-network model, taking historical operating data, building usage plans, and historical deviation values ​​as input. The optimization objective is to ensure that the operating effect deviation value accurately represents the degree of inaccuracy, outputting the adjustment amount of the initial weights to achieve weight optimization.

[0068] Specifically, the deep Q-network model used for adaptive weight adjustment adopts a three-level hierarchical structure: state input layer, feature extraction layer, and Q-value output layer. The layers are mapped from state to action value through fully connected layers and nonlinear transformations. Each layer is closely related and logically clear: the state input layer, as the model's information receiving unit, has a number of neurons matching the dimension of the input state, including historical energy consumption deviation indicators, economic deviation indicators, temperature deviation indicators, and building usage plan data from multiple control cycles. It is responsible for converting multi-dimensional state data into a vector form that the model can process. The feature extraction layer is the core processing unit of the model, with two fully connected hidden layers. It uses the ReLU activation function to mine latent features related to weight adjustment in the state data, achieving a mapping from the original state to a high-dimensional feature space. The Q-value output layer is the model's decision output unit, with a number of neurons consistent with the dimension of the selectable action space for weight adjustment. It is responsible for outputting the Q-value (action value) corresponding to each weight adjustment action, providing a decision basis for weight coefficient optimization. The three-layer structure forms a complete link of state perception, feature mining, and action estimation, ensuring the adaptability and accuracy of weight adjustment.

[0069] In this embodiment, the weight adjustment amount is a correction value (positive or negative) output by the reinforcement learning model to the initial weight coefficient set. It is used to fine-tune the initial weights so that the final weights better match the operating state of the target building. The final weight coefficient is the weight value obtained by superimposing the initial weight coefficient set and the weight adjustment amount. It satisfies the condition that the sum of the energy consumption deviation weight, economic deviation weight, and temperature deviation weight is 1, and is the final basis for calculating the operating effect deviation value.

[0070] As can be seen from the above, this embodiment can obtain final weight coefficients that are adapted to different buildings and operating stages by matching the initial weight coefficient set according to the target building attributes and operating season, and adjusting the initial weight coefficient set through a reinforcement learning weight adjustment model. Using these coefficients to calculate the deviation value of the operating effect, the deviation between the operating effect and the prediction of the building air conditioning system can be represented more accurately, providing a more reliable basis for calibration and optimization.

[0071] In one embodiment of this application, the determination of the weighting coefficients further includes: Construct a multi-objective optimization problem with the objectives of minimizing energy consumption deviation, minimizing economic deviation, and minimizing temperature deviation. An online multi-objective evolutionary algorithm is used to iteratively solve a set of non-dominated solutions using historical actual running data as samples. From the set of non-dominated solutions, select a target solution according to the current operating strategy mode, and use the weight coefficient corresponding to the target solution as the final weight coefficient.

[0072] In this embodiment, the multi-objective optimization problem is a mathematical optimization problem with the objectives of minimizing energy consumption deviation, minimizing economic deviation, and minimizing temperature deviation. These three objectives are interdependent; for example, reducing energy consumption deviation may lead to an increase in temperature deviation. It is impossible to achieve the optimal solution simultaneously across all objectives, necessitating a balance among them. Online multi-objective evolutionary algorithms are a class of optimization algorithms based on biological evolutionary mechanisms (selection, crossover, and mutation). They can directly handle multi-objective optimization problems and support online iterative solutions using real-time historical data as samples, resulting in diverse output solutions.

[0073] Specifically, the online multi-objective evolutionary algorithm used to solve for the non-dominated solution set of weight coefficients adopts the NSGA-II framework (Non-dominated sorting genetic algorithm II). It employs a three-level closed-loop hierarchical structure: online data interaction layer, population evolution layer, and solution set selection layer. The functions and relationships of each layer are adapted to the optimization scenario of building air conditioning operation parameters: The online data interaction layer, as the front-end input and output unit of the algorithm, is responsible for collecting historical actual operation data (energy consumption, electricity cost, indoor temperature) and deviation index calculation results of the building air conditioning system, and feeding back the iteratively generated weight coefficient solution set to the optimization control model for verification; The population evolution layer is the core calculation unit of the algorithm, including sub-modules such as initialization, selection, crossover, mutation, non-dominated sorting, and crowding calculation, and achieves global optimization of weight coefficients through multiple generations of population iteration; The solution set selection layer is the back-end decision-making unit of the algorithm. Based on the current operation strategy mode of the building (economy priority / energy efficiency priority / comfort priority), it selects the optimal weight combination that fits the business needs from the candidate solution set obtained by evolution. The three-layer structure forms an online iterative link of data input-evolution optimization-solution set selection and verification, ensuring the practicality and adaptability of the solution set.

[0074] In this embodiment, the non-dominated solution set is the optimal solution set for a multi-objective optimization problem. No solution in this set is dominated by any other solution; that is, it is impossible to improve one objective without degrading the performance of others. This solution set provides multiple candidate solutions with different objective priorities. The operating strategy mode is the preset operating priority mode for the target building, including economy priority mode, energy efficiency priority mode, comfort priority mode, and equilibrium mode. It serves as the basis for selecting target solutions from the non-dominated solution set. The target solution is the solution selected from the non-dominated solution set based on the current operating strategy mode that best fits the actual operational needs. The weight coefficient corresponding to this solution is the weight ultimately used to calculate the deviation value of the operating effect.

[0075] As can be seen from the above, this embodiment constructs a multi-objective optimization problem and uses an online multi-objective evolutionary algorithm to iteratively solve it to obtain a non-dominated solution set. It can determine the weight coefficients by comprehensively considering energy consumption, economy, and temperature deviation indicators based on historical actual operating data. From the non-dominated solution set, the target solution is selected as the final weight coefficient according to the current operating strategy mode, so that the weight coefficients are more in line with the actual operating strategy, thereby more accurately calculating the operating effect deviation value and providing a more reasonable decision-making basis for the energy-saving operation simulation and optimization of building air conditioning systems.

[0076] Corresponding to the energy-saving operation simulation and optimization method of building air conditioning system in the above embodiment, Figure 2 This is a structural block diagram of a building air conditioning system energy-saving operation simulation and optimization system provided in one embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. References Figure 2 The building air conditioning system energy-saving operation simulation and optimization system 20 includes: model construction module 21, simulation prediction module 22, model establishment module 23, optimization solution module 24, operation deviation module 25, decision update module 26, and model application module 27.

[0077] Among them, the model building module 21 is used to acquire multi-source data of the target building, and to build and initialize the building air conditioning simulation model based on the multi-source data; The simulation prediction module 22 is used to acquire meteorological forecast data and building usage plan data for the next control cycle, input them into the building air conditioning simulation model for forward simulation, and obtain the predicted operation data corresponding to the next control cycle. Model module 23 is established to construct a comprehensive objective function that minimizes total operating energy consumption and total operating cost. An optimization control model is established with indoor environmental parameters in the comfort range and equipment operating parameters in the safe range as constraints. The optimization solution module 24 is used to solve the optimization control model using optimization algorithms to obtain the optimal control command sequence for the equipment in the next control cycle; The operation deviation module 25 is used to execute the optimal control command sequence, collect actual operation data, calculate the deviation between the actual operation data and the predicted operation data, and obtain the operation effect deviation value. The decision update module 26 is used to determine, based on historical actual operating data of multiple control cycles and preset decision logic, to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model if the deviation value of the operating effect is greater than the preset deviation threshold. Model application module 27 is used to apply the calibrated building air conditioning simulation model and / or the updated optimized control model to subsequent control cycles.

[0078] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the modules in the aforementioned device embodiments, for example... Figure 2 The functions of the model building module 21, simulation prediction module 22, model building module 23, optimization solution module 24, operation deviation module 25, decision update module 26, and model application module 27 are shown.

[0079] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0080] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0081] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store device type information.

[0082] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation methods described in any embodiment of the energy-saving operation simulation and optimization method for building air conditioning systems provided in the embodiments of this application, or they can execute the implementation methods of the electronic devices described in the embodiments of this application, which will not be repeated here.

[0083] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0084] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0085] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0086] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0087] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces or units, or it may be an electrical, mechanical, or other form of connection.

[0088] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0089] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0090] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for energy-saving operation simulation and optimization of building air conditioning systems, characterized in that, include: Acquire multi-source data of the target building, and construct and initialize a building air conditioning simulation model based on the multi-source data; Obtain meteorological forecast data and building usage plan data for the next control cycle, input them into the building air conditioning simulation model for forward simulation, and obtain the predicted operating data corresponding to the next control cycle; A comprehensive objective function is constructed to minimize total operating energy consumption and total operating cost. An optimal control model is established with indoor environmental parameters in the comfort range and equipment operating parameters in the safe range as constraints. The optimal control model is solved using an optimization algorithm to obtain the optimal control command sequence for the equipment in the next control cycle. Execute the optimal control command sequence, collect actual operating data, calculate the deviation between the actual operating data and the predicted operating data, and obtain the operating effect deviation value; If the deviation value of the operating effect is greater than the preset deviation threshold, then based on the historical actual operating data of multiple control cycles and the preset decision logic, it is decided to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model. The calibrated building air conditioning simulation model and / or the updated optimized control model are applied to subsequent control cycles.

2. The method for energy-saving operation simulation and optimization of a building air conditioning system according to claim 1, characterized in that, The step of calculating the deviation between the actual operating data and the predicted operating data to obtain the operating performance deviation value includes: The relative error between the actual total energy consumption in the actual energy consumption curve and the corresponding predicted total energy consumption in the predicted energy consumption curve is calculated and used as an energy consumption deviation index. The actual total electricity cost is calculated based on the time-of-use electricity price data and the actual energy consumption curve. The predicted total electricity cost is calculated based on the energy consumption prediction curve and time-of-use electricity price data. The relative error between the actual total electricity cost and the predicted total electricity cost is calculated as an economic deviation indicator; The average absolute error between the average actual indoor temperature and the predicted indoor temperature of the target building is calculated and used as a temperature deviation index. The weighted sum of the energy consumption deviation index, the economic deviation index, and the temperature deviation index is taken as the operational performance deviation value.

3. The method for energy-saving operation simulation and optimization of a building air conditioning system according to claim 1, characterized in that, If the operational performance deviation exceeds a preset deviation threshold, then based on historical operational data from multiple control cycles and preset decision logic, it is determined to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model, including: Obtain the operational performance deviation values ​​for the current and historical control cycles, along with their corresponding energy consumption deviation, economic performance deviation, and temperature deviation indicators. The analysis includes the number of cycles in which the operational performance deviation value exceeds the preset deviation threshold, the slope of the trend, and the contribution of each of the energy consumption deviation index, the economic deviation index, and the temperature deviation index to the total deviation. If the average contribution ratio of the energy consumption deviation index is greater than the contribution threshold for three consecutive control cycles and the slope of its change trend is greater than the preset slope threshold, then the building air conditioning simulation model is calibrated. If the average contribution ratio of the economic deviation index is greater than the contribution threshold for three consecutive control cycles, then the weight coefficient of the comprehensive objective function and the boundary parameters of the constraint conditions in the optimized control model will be adjusted. If the temperature deviation index exceeds 1.5 times its preset deviation threshold for two consecutive control cycles, the room temperature prediction module of the building air conditioning simulation model will be calibrated and the comfort range constraints in the optimized control model will be reviewed and updated.

4. The method for energy-saving operation simulation and optimization of a building air conditioning system according to claim 3, characterized in that, The calibration of the building air conditioning simulation model includes: By comparing and analyzing the deviation between the predicted energy consumption curve and the actual energy consumption curve, the error sources causing the deviation are identified; the error sources include inaccurate building thermal parameters or inaccurate equipment performance models. Based on the error sources, a calibration dataset is constructed from historical operational data, and calibration operations are performed: If the error source is inaccurate building thermal parameters, then the parameters of the building air conditioning simulation model are updated by inversion using an optimization algorithm; If the error source is an inaccurate equipment performance model, then the historical actual operating data in the calibration dataset is used as the training target, and the neural network is used to correct the equipment performance mapping relationship in the original building air conditioning simulation model.

5. The method for energy-saving operation simulation and optimization of a building air conditioning system according to claim 3, characterized in that, The adjustment of the weight coefficients of the integrated objective function and the boundary parameters of the constraints in the optimized control model includes: Multidimensional regression analysis was performed on the data of the aforementioned economic deviation indicators to identify the sources of deviation. If the source of deviation is that the boundary parameter settings of the constraint conditions do not match the actual operating range, then the boundary parameters in the constraint conditions should be corrected. If the source of deviation is that the weight coefficients of the two terms in the comprehensive objective function, namely minimizing total operating energy consumption and minimizing total operating cost, do not match the priority of actual business needs, then the weight coefficients of the corresponding objective terms in the comprehensive objective function shall be adjusted.

6. The method for energy-saving operation simulation and optimization of a building air conditioning system according to claim 5, characterized in that, If the source of deviation is that the boundary parameter settings of the constraint conditions do not match the actual operating range, then the boundary parameters in the constraint conditions are corrected, including: Based on the multidimensional regression analysis, constraints that are related to the economic deviation index and have become active or tight constraints in historical optimization solutions are identified as constraints to be corrected. Extract the data sequence of operating parameters corresponding to the constraints to be corrected from the historical actual operating data; Based on the statistical distribution of the data sequence, a preset confidence quantile interval is calculated as a candidate boundary value; The intersection of the candidate boundary values ​​and the allowable parameter range of the building air conditioning system design code is used as the safety boundary parameter after the constraint conditions are updated.

7. The method for energy-saving operation simulation and optimization of a building air conditioning system according to claim 5, characterized in that, If the source of deviation is a mismatch between the weight coefficients of the two terms in the comprehensive objective function (minimizing total operating energy consumption and minimizing total operating cost) and the priority of actual business needs, then the weight coefficients of the corresponding objective terms in the comprehensive objective function shall be adjusted, including: Obtain the priority factor of the actual business needs, the priority being determined by querying a preset operation strategy mapping table; Based on the actual operating data, calculate the performance index deviation rate of the total operating energy consumption item and the total operating cost item. The deviation rate includes the energy consumption efficiency index deviation rate and the operating cost control index deviation rate. If the deviation rate of the energy consumption efficiency index is greater than the deviation rate of the operating cost control index, and the priority factor indicates that energy efficiency is prioritized, then the weight of the target of minimizing total energy consumption is increased based on the first adjustment step size, and the weight of the target of minimizing total system operating cost is decreased based on the first adjustment step size. If the deviation rate of the operating cost control index is greater than the deviation rate of the energy consumption efficiency index, and the priority factor indicates that economic efficiency takes precedence, then the weight of the target of minimizing the total operating cost of the system is increased based on the second adjustment step size, and the weight of the target of minimizing the total energy consumption is decreased based on the second adjustment step size.

8. A simulation and optimization system for energy-saving operation of a building air conditioning system, characterized in that, include: A model building module is used to acquire multi-source data of the target building, and to build and initialize a building air conditioning simulation model based on the multi-source data; The simulation prediction module is used to acquire meteorological forecast data and building usage plan data for the next control cycle, input them into the building air conditioning simulation model for forward simulation, and obtain the predicted operation data corresponding to the next control cycle. A model module is established to construct a comprehensive objective function that minimizes total operating energy consumption and total operating cost. An optimization control model is established with indoor environmental parameters in the comfort range and equipment operating parameters in the safe range as constraints. The optimization solution module is used to solve the optimized control model using optimization algorithms to obtain the optimal control command sequence for the equipment in the next control cycle; The operation deviation module is used to execute the optimal control command sequence, collect actual operation data, calculate the deviation between the actual operation data and the predicted operation data, and obtain the operation effect deviation value. The decision update module is used to determine, based on historical actual operating data of multiple control cycles and preset decision logic, to calibrate the building air conditioning simulation model and / or iteratively update the optimized control model if the deviation value of the operating effect is greater than a preset deviation threshold. The model application module is used to apply the calibrated building air conditioning simulation model and / or the updated optimized control model to subsequent control cycles.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images