Energy consumption grading optimization method, computer equipment and storage medium

Abstract

The invention relates to an energy consumption hierarchical optimization method, computer equipment and a storage medium, and relates to the field of energy. The method comprises the steps of obtaining target operation data related to energy consumption optimization; performing data classification on the target operation data to obtain a data set corresponding to each optimization hierarchy; the data sets of different optimization levels correspond to different optimization indexes; determining an initial optimization level based on the equipment operation mechanism and the current state indicated by the target operation data; taking the initial optimization hierarchy as an optimization starting point, performing hierarchical optimization calculation based on the optimization model of each optimization hierarchy and the data set corresponding to each optimization hierarchy to obtain a target energy consumption optimization strategy, and performing energy consumption optimization on the heating and ventilation equipment based on the target energy consumption optimization strategy; through the method, the optimization efficiency can be improved, and the practicability of the energy consumption optimization strategy can be improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of energy, and in particular to an energy consumption classification optimization method, a computer device, and a storage medium. Background Art

[0002] Reducing carbon emissions is an important means of environmental protection. In some scenarios where heating equipment is used (such as data centers), energy consumption can be reduced by optimizing the control strategy of HVAC equipment to reduce carbon emissions.

[0003] In related technologies, the method for optimizing the control strategy of HVAC equipment is mainly to directly establish an energy consumption prediction model between the operating data related to energy consumption optimization and the electricity utilization efficiency through machine learning or deep learning. The operating data related to energy consumption optimization includes outdoor environmental data, HVAC equipment, and key measurement point data of IT equipment. If the predicted electricity utilization efficiency is lower than the original electricity utilization efficiency, the input parameter combination of the model is considered to be an effective energy consumption optimization method.

[0004] However, the above optimization method is a purely data-driven model. In theory, it can achieve good energy consumption optimization effects. However, in practice, it may cause frequent start and stop of equipment due to frequent adjustment of control parameters or large parameter differences, which may be inconsistent with the actual operation and maintenance logic, making the energy consumption optimization method less practical. Summary of the invention

[0005] The embodiment of the present application provides an energy consumption classification optimization method, a computer device and a storage medium, which can improve the optimization efficiency and the practicality of the energy consumption optimization strategy. The technical solution is as follows:

[0006] In one aspect, a method for optimizing energy consumption by grading is provided, the method comprising:

[0007] Obtain target operating data related to energy consumption optimization;

[0008] The target operation data is classified to obtain data sets corresponding to various optimization levels; data sets at different optimization levels correspond to different optimization indicators;

[0009] Determining a starting optimization level based on the equipment operation mechanism and the current state indicated by the target operation data;

[0010] Taking the initial optimization level as the starting point for optimization, hierarchical optimization calculations are performed based on the optimization models of each optimization level and the data sets corresponding to each optimization level to obtain a target energy consumption optimization strategy, so as to optimize the energy consumption of HVAC equipment based on the target energy consumption optimization strategy.

[0011] On the other hand, a computer device is provided, which includes a processor and a memory, wherein the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the above-mentioned energy consumption classification optimization method.

[0012] On the other hand, a readable storage medium is provided, in which at least one computer program is stored. The computer program is loaded and executed by a processor to implement the above-mentioned energy consumption classification optimization method.

[0013] On the other hand, a computer program product is provided, the computer program product comprising at least one computer program, the computer program being loaded and executed by a processor to implement the energy consumption classification optimization method provided in the above-mentioned various optional implementations.

[0014] The technical solution provided by this application may have the following beneficial effects:

[0015] The energy consumption hierarchical optimization method provided in the embodiment of the present application obtains target operating data related to energy consumption optimization, performs data grading on the target operating data to obtain data sets of each optimization level corresponding to different optimization indicators, and after the starting optimization level is determined according to the operation mechanism of the equipment and the current state, takes the starting optimization level in each optimization level as the starting point for optimization, performs hierarchical optimization calculations based on the optimization models of each optimization level and the data sets of each optimization level to obtain a target energy consumption optimization strategy that meets the current actual needs, and optimizes the energy consumption of the HVAC equipment based on the target energy consumption optimization strategy; through the above method, the starting optimization level that is compatible with the equipment operation mechanism and the current state can be determined, and the energy consumption optimization strategy can be formulated by performing hierarchical optimization based on the starting optimization level as the starting point for hierarchical optimization, which can avoid unnecessary optimization processes and improve optimization efficiency. In addition, since the obtained target energy consumption optimization strategy is compatible with the current state and the equipment operation mechanism, the practicability of the energy consumption optimization strategy is improved.

[0016] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0018] Figure 1 A flow chart of an energy consumption classification optimization method provided by an exemplary embodiment of the present application is shown;

[0019] Figure 2 A flow chart of an energy consumption classification optimization method provided by an exemplary embodiment of the present application is shown;

[0020] Figure 3 A flow chart of an energy consumption classification optimization method provided by an exemplary embodiment of the present application is shown;

[0021] Figure 4 A schematic diagram of an energy consumption classification optimization system provided by an exemplary embodiment of the present application is shown;

[0022] Figure 5 A structural block diagram of a computer device shown in an exemplary embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0024] In the scenario where HVAC equipment is installed, energy saving and energy reduction can be achieved by optimizing the energy consumption of HVAC equipment. Schematically, energy saving and energy reduction can be achieved by optimizing the energy consumption of HVAC equipment. At present, the formulation of HVAC equipment energy consumption optimization strategy has the following problems:

[0025] 1) The relationship between the control parameters of HVAC equipment is not considered, and all control parameters are treated equally (such as the number of equipment in operation and the internal control parameters of the equipment are adjusted equally during each optimization). The changes in the output control parameters (especially the difference in the number of equipment in operation) will cause the equipment to start and stop frequently, which is not in line with the actual operation and maintenance of the machine.

[0026] 2) It does not take into account that the combination of HVAC equipment types (i.e., operating modes) will vary according to climate change, and the corresponding internal control parameters of the equipment will also be different. As a result, the internal control parameters given by the existing model are very different from the parameter setting logic of the current actual operation and maintenance of the existing equipment, which makes it difficult to be credibly accepted.

[0027] 3) Safety and effectiveness were not fully considered in the modeling process in combination with actual operation and maintenance, resulting in low credibility of the optimization control strategy output by the model when it was subsequently issued.

[0028] 4) The huge difference between the actual collected messy data and the ideal simulated data is not taken into account. The actual collected data has a lot of problems such as missing data, inconsistency between equipment status indicators and equipment frequency feedback, and data collection anomalies. It is necessary to combine the equipment operation mechanism for hierarchical processing and correctness verification. If effective data grading is not performed, the model accuracy will be low and subsequent applications will be invalid.

[0029] In view of the above problems, the embodiments of the present application provide an energy consumption classification optimization method, which can solve all or part of the above problems, and improve the practicality of the energy consumption optimization strategy while optimizing energy consumption. The energy consumption classification optimization method of the present application is introduced through several embodiments below.

[0030] Figure 1 A flowchart of an energy consumption classification optimization method provided by an exemplary embodiment of the present application is shown. The method can be executed by a computer device, and the computer device can be implemented as a management device for HVAC equipment, such as Figure 1 As shown, the energy consumption classification optimization method may include the following steps:

[0031] Step 110, obtaining target operation data related to energy consumption optimization.

[0032] In the embodiment of the present application, the target operation data related to energy consumption optimization may include temperature and humidity of various links indoors and outdoors, energy consumption of various HVAC equipment, status information of key components of HVAC equipment, and control parameters, etc. The target operation data may be obtained by computer equipment monitoring various HVAC equipment.

[0033] Step 120 , classify the target operation data to obtain data sets corresponding to various optimization levels; data sets at different optimization levels correspond to different optimization indicators.

[0034] When formulating an energy consumption optimization strategy based on target operating data, the target operating data must first be classified according to the optimization indicators corresponding to each optimization level to obtain a data set corresponding to each optimization level; different optimization levels correspond to different optimization indicators, which is reflected in the number of optimization indicators involved in different optimization levels and the different types of optimization indicators. The optimization indicators involved in different optimization levels may be partially the same or completely different; schematically, one optimization level may involve optimization indicator A, and another optimization level may involve optimization indicator A and optimization indicator B. The number of optimization levels and the settings of the optimization indicators involved in each optimization level can be set by relevant personnel based on actual needs, and this application does not impose any restrictions on this.

[0035] Step 130, determining a starting optimization level based on the equipment operation mechanism and the current state indicated by the target operation data.

[0036] In the embodiment of the present application, each optimization level may have a hierarchical relationship, that is, in the process of formulating the energy consumption optimization strategy, the computer device may perform optimization calculation step by step according to the hierarchical relationship between each optimization level. Therefore, in order to avoid meaningless optimization calculation process, the computer device may first determine the starting optimization level in the current state according to the device operation mechanism and the current state indicated by the target operation data, so as to determine the optimization starting point in each optimization level.

[0037] Step 140, taking the initial optimization level as the starting point for optimization, performing hierarchical optimization calculations based on the optimization models of each optimization level and the data sets corresponding to each optimization level, and obtaining a target energy consumption optimization strategy to optimize the energy consumption of HVAC equipment based on the target energy consumption optimization strategy.

[0038] In schematic form, if there are four hierarchical optimization levels, the starting optimization level determined based on the equipment operation mechanism and the current state indicated by the target operation data is the second optimization level. When formulating the energy consumption optimization strategy, the first optimization level can be skipped, and hierarchical optimization calculation and verification can be performed from the second optimization level until the target energy consumption optimization strategy is obtained or all optimization levels are traversed. The formulation of energy consumption optimization strategies based on different optimization indicators can be realized through hierarchical optimization. By determining the starting optimization layer, unnecessary optimization calculation and verification processes can be avoided, thereby improving the efficiency of formulating energy consumption optimization strategies and improving the applicability of energy consumption optimization strategies to the current state.

[0039] When each optimization level is performing optimization calculations, the optimization model of the optimization level is used to perform optimization calculations on the data set of the optimization level to obtain the optimization result of the optimization level, that is, the energy consumption optimization strategy of the optimization level; in one possible implementation method, the computer device can directly determine the energy consumption optimization strategy obtained by the optimization level as the target energy consumption optimization strategy, or, in another possible implementation method, the computer device can determine the energy consumption optimization strategy as the target energy consumption optimization strategy when the energy consumption optimization strategy meets specified conditions.

[0040] In summary, the energy consumption hierarchical optimization method provided in the embodiment of the present application obtains target operating data related to energy consumption optimization, performs data grading on the target operating data to obtain data sets of each optimization level corresponding to different optimization indicators, and after the starting optimization level is determined according to the operation mechanism of the equipment and the current state, the starting optimization level in each optimization level is used as the starting point for optimization, and a hierarchical optimization calculation is performed based on the optimization model of each optimization level and the data set of each optimization level to obtain a target energy consumption optimization strategy that meets the current actual needs, and the energy consumption of the HVAC equipment is optimized based on the target energy consumption optimization strategy; through the above method, the starting optimization level that is compatible with the equipment operation mechanism and the current state can be determined, and an energy consumption optimization strategy can be formulated by performing hierarchical optimization based on the starting optimization level as the starting point for hierarchical optimization, which can avoid unnecessary optimization processes and improve optimization efficiency. In addition, since the obtained target energy consumption optimization strategy is compatible with the current state and the equipment operation mechanism, the practicality of the energy consumption optimization strategy is improved.

[0041] In a possible implementation, when the energy consumption optimization strategy output by the optimization level passes the security and effectiveness verification, the energy consumption optimization strategy is determined as the target energy consumption optimization strategy; in the embodiment of the present application, the computer device can implement the security and effectiveness verification of the energy consumption optimization strategy through the security and effectiveness verification model; in this case, Figure 2 A flowchart of an energy consumption classification optimization method provided by an exemplary embodiment of the present application is shown. The method can be executed by a computer device, and the computer device can be implemented as a management device for HVAC equipment, such as Figure 2 As shown, the energy consumption classification optimization method may include the following steps:

[0042] Step 201, obtaining target operation data related to energy consumption optimization.

[0043] The computer device can obtain the real operation data obtained from the real data center as the target operation data, and the real data center may include an IT center and a supervision system (SS). In the embodiment of the present application, the target operation data may include indoor and outdoor temperature and humidity and information technology IT (Information Technology) load changes; the target operation data is used to indicate the current state.

[0044] Step 202 , classify the target operation data to obtain data sets corresponding to various optimization levels; data sets at different optimization levels correspond to different optimization indicators.

[0045] The computer device can implement a process of performing data classification on the target operation data through the data classification processing module to obtain data sets corresponding to each optimization level, and the process may include:

[0046] The target operation data is subjected to data classification cleaning, data classification verification, index classification fusion, index classification extraction, and index classification division to obtain a data set applied to the control strategy classification optimization model; the control strategy classification optimization model includes optimization models for each optimization level; the data set applied to the control strategy classification optimization model includes data subsets of the optimization models applied to each optimization level.

[0047] In the data classification process, data classification cleaning is used to clean redundant data in the target running data, such as checking data consistency, processing invalid values ​​and missing values, etc.

[0048] Data hierarchical verification is used to verify the correctness of each data; data hierarchical verification can verify whether there are abnormalities in the data collection of each device based on the business logic. When it is determined that there are abnormalities, it is determined whether this part of the abnormality belongs to equipment failure or data collection error, and the size of this part of abnormal data is counted. On this basis, corresponding data correction or data removal measures can be taken based on the pre-specified data correctness extraction logic; schematically, when the business logic based on the data hierarchical verification is the business logic of the chilled water refrigeration system, it is necessary to determine whether the cooling pump, the primary refrigeration pump, the water chiller or the plate changer in the chiller meet the conditions of being turned on and off at the same time, so as to determine whether there are abnormalities in the data collection of each device corresponding to the chilled water refrigeration system. For example, when the device status bit of a certain device shows that the device is turned off, but the frequency or current percentage of the device has a value showing that it is turned on, it is determined that there is an abnormality.

[0049] Indicator hierarchical fusion is used to collect data statistics on related indicators and generate new indicators. Schematically, computer equipment determines the actual startup status of the equipment (i.e., new indicators) through the equipment status indicators, control indicators, and control feedback indicators. In addition, indicator hierarchical fusion can also be implemented to count the number of running devices of the same type, generate temperature and humidity averages, generate current percentage averages, and generate frequency feedback averages.

[0050] Indicator hierarchical extraction is used to extract indicator sets required by different models to form corresponding data subsets. Schematically, the data subset Data_E applied to the energy consumption effectiveness verification model contains indicators related to energy consumption, such as indicators related to PUE (Power Usage Effectiveness) or CLF (Cooling Load Factor), where PUE is the ratio of all energy consumed to the energy used by the IT load, and CLF is the ratio of the power consumption of the refrigeration equipment to the power consumption of the IT equipment; the data subset Data_C applied to the cooling capacity safety verification model contains indicators related to the cooling capacity Cooling; the data subset Data_S applied to the state safety verification contains indicators related to the temperature and pressure state; the data subsets of each optimization level required by the control strategy hierarchical optimization model contain indicators related to the current optimization level.

[0051] Indicator classification is used to classify each indicator. Schematically, optimization indicators can be divided into different types based on indicator type, indicator function and other classification criteria. For example, based on the indicator type, each indicator can be divided into environmental indicators, equipment operation status indicators, equipment control feedback indicators, equipment control parameter setting indicators, various energy consumption indicators, etc.; based on the indicator function, each indicator can be divided into safety verification-related indicators, validity verification-related indicators and strategy classification optimization-related indicators, etc. By performing indicator classification, the iteration and correction of the data classification processing process can be facilitated.

[0052] Optionally, before classifying the target operation data to obtain data sets corresponding to each optimization level, the method further includes:

[0053] Performing optimization pre-detection on the target operation data through optimization pre-detection indicators to obtain optimization pre-detection results; the optimization pre-detection indicators include at least one of the following: checking whether the indicators required by each model are missing and whether the business relationship between the indicators is established;

[0054] When the optimization pre-detection result indicates that the target operation data has passed the pre-detection, the target operation data is classified to obtain data sets corresponding to each optimization level.

[0055] That is to say, when entering the online reasoning process of energy consumption classification optimization, the target operation data will be pre-checked for optimization first. After the target operation data passes the pre-check for optimization, the subsequent processing process will be carried out. Among them, checking whether the indicators required by each model are missing can be to check whether there are large areas of missing indoor and outdoor ambient temperatures, equipment operating status, equipment control feedback, equipment setting parameters, etc.; checking whether the business relationship between indicators is established can be to check whether the status of each device in the chiller is in compliance with the same opening and closing, etc.

[0056] When the optimization pre-detection result indicates that the target operation data detection fails, it is determined that the HVAC equipment is not suitable for energy consumption optimization.

[0057] When the optimization pre-detection results indicate that the target operation data detection fails, it means that there may be equipment failure or problems with data collection. In this case, it is determined that the HVAC equipment is not suitable for energy consumption optimization and the subsequent process is temporarily not entered; optionally, in this case, a "not suitable for optimization" prompt can also be given through audio information or visual information, such as issuing a prompt sound or displaying text or image prompt information on a display device, etc.

[0058] Step 203, determining the initial optimization level based on the equipment operation mechanism and the current state indicated by the target operation data.

[0059] Schematically, take the example of three optimization levels configured in a computer device, namely the first optimization level, the second optimization level and the third optimization level, and there is a hierarchical relationship between the three optimization levels. When performing hierarchical optimization calculation, if the starting optimization level is the first optimization level, the optimization is performed in the order from the first optimization level to the third optimization level; the optimization indicators of the three optimization levels are different, among which the first optimization level refers to optimizing the internal control parameters of the HVAC equipment; the second optimization level refers to optimizing the operating quantity and internal control parameters of the HVAC equipment; the third optimization level refers to optimizing the operating mode, operating quantity and internal control parameters of the HVAC equipment; wherein the operating mode reflects the combination of the equipment, for example, the chilled water refrigeration equipment is divided into free mode, pre-cooling mode, refrigeration mode, etc. Based on this, the embodiment of the present application provides a feasible method for determining the starting optimization level based on climate change, and the process of determining the starting optimization level by the computer device can be implemented as follows:

[0060] When the outdoor temperature is within the preset outdoor temperature range, the indoor and outdoor temperature and humidity change periodically, and the IT load change periodically, the starting optimization level is determined to be the first optimization level; the first optimization level indicates that the internal control parameters of the HVAC equipment are optimized;

[0061] When the outdoor temperature is within the outdoor temperature range, the indoor and outdoor temperature and humidity changes indicate that there is a sudden change in the indoor and outdoor temperature and humidity, or the IT load changes indicate that there is a sudden change in the IT load, the starting optimization level is determined to be the second optimization level; the second optimization level indicates that the operating quantity and internal control parameters of the HVAC equipment are optimized;

[0062] When the outdoor temperature is within the critical temperature range between different outdoor temperature ranges, the starting optimization level is determined to be the third optimization level; the third optimization level indicates optimizing the operation mode, operation quantity and internal control parameters of the HVAC equipment.

[0063] That is to say, the optimization indicators of the first optimization level are the internal control parameters of the HVAC equipment, the optimization indicators of the second optimization level are the operating quantity and internal control parameters of the HVAC equipment, and the optimization indicators of the third optimization level are the operating mode, operating quantity and internal control parameters of the HVAC equipment.

[0064] When the outdoor temperature is within the preset outdoor temperature range, it indicates that the climate has not changed significantly. The preset outdoor temperature range can be set based on the season. For example, the outdoor temperature range in winter can be T≤10, the outdoor temperature range in spring and autumn can be 10<T≤26, and the outdoor temperature range in summer can be T>26. It should be noted that the above setting of the outdoor temperature range is only for illustration. Based on different actual conditions, the outdoor temperature range can also have different settings, and this application does not limit this. When the climate has not changed significantly, and the indoor and outdoor temperature and humidity and IT load are in periodic changes, when optimizing energy consumption, it is preferred to not adjust the operating mode and the number of equipment in operation of the HVAC equipment, but to optimize the internal control parameters of the HVAC equipment, that is, to perform an optimization calculation process with the first optimization level as the initial optimization level.

[0065] When the outdoor temperature is within the outdoor temperature range, that is, the climate has not changed significantly, the changes in indoor and outdoor temperature and humidity indicate that there is a sudden change in the indoor and outdoor temperature and humidity, or the changes in IT load indicate that there is a sudden change in the IT load, when optimizing energy consumption, it is preferred to not adjust the operating mode of the HVAC equipment, but to optimize the operating quantity of the HVAC equipment and the internal control parameters of the equipment, that is, to execute the optimization calculation process with the second optimization level as the initial optimization level; optionally, the computer device can execute the optimization calculation process with the second optimization level as the initial optimization level when the climate has not changed significantly, the indoor and outdoor temperature and humidity have a sudden change and exceed the specified proportion threshold within a specified time interval, or the IT load has a sudden change and exceeds the specified proportion threshold within a period of time; for example, if the outdoor temperature is within the outdoor temperature range and the IT load has a sudden change and the change within 4 hours exceeds 50%, it is determined to execute the optimization calculation process with the second optimization level as the initial optimization level; the specified time interval and the specified proportion threshold can be set based on actual needs, and this application does not limit this.

[0066] When the outdoor temperature is within the critical temperature range between different outdoor temperature ranges, it indicates that the climate is in a season transition state. For example, corresponding to the outdoor temperature range based on season settings, when the outdoor temperature T is around 10 or 26, the climate is determined to be in a season transition state. At this time, the optimization method of optimizing the operating mode of the HVAC equipment, the operating number of the HVAC equipment, and the internal control parameters of the equipment is preferentially selected, that is, executing the optimization calculation process with the third optimization level as the initial optimization level.

[0067] It should be noted that the above-mentioned division of optimization levels and the setting of climate conditions when each optimization level is used as the initial optimization level are only a feasible division and setting method provided in the embodiment of the present application. Relevant personnel can set more or fewer optimization levels and set corresponding climate conditions based on actual needs, and the present application does not impose any restrictions on this.

[0068] Step 204, when entering the target optimization level, use the optimization model of the target optimization level to perform optimization calculations on the corresponding data set to obtain a candidate energy consumption optimization strategy; the target optimization level is any one of the optimization levels that uses the initial optimization level as the optimization starting point for optimization.

[0069] In the hierarchical optimization process, the computer device starts the optimization process with the initial optimization level as the starting point to obtain the target energy consumption optimization strategy or ends the optimization process after traversing each optimization level; in the process of hierarchical optimization with the determined starting optimization level as the starting point, in order to further improve the practicality of the obtained energy consumption optimization strategy, after the energy consumption optimization strategy is obtained at each optimization level in the optimization process, a safety and effectiveness verification process will be carried out to determine whether it is determined as the target energy consumption optimization strategy based on the verification results.

[0070] Step 205 , performing safety and effectiveness verification on the candidate energy consumption optimization strategy through the safety and effectiveness verification model to obtain the verification result of the candidate energy consumption optimization strategy.

[0071] In an embodiment of the present application, the safety and effectiveness verification of the energy consumption optimization strategy performed by the computer device may include: energy consumption effectiveness verification, cooling capacity safety verification and application status security verification; correspondingly, the safety and effectiveness verification model may include: energy consumption effectiveness verification model, cooling capacity safety verification model and application status security verification model.

[0072] Among them, the energy consumption optimization strategy obtained by the optimization model is verified for safety and effectiveness through the safety and effectiveness verification model, including:

[0073] The energy efficiency verification model is used to verify the energy efficiency of the energy optimization strategy, and a first sub-verification result is obtained;

[0074] The cooling capacity safety verification model is used to verify the cooling capacity safety of the energy consumption optimization strategy, and the second sub-verification result is obtained;

[0075] The energy consumption optimization strategy is verified for state safety through the state safety verification model to obtain the third sub-verification result;

[0076] Based on the first sub-verification result, the second sub-verification result and the third sub-verification result, a verification result is generated; the verification result indicates whether the energy consumption optimization strategy has passed the verification or has failed the verification.

[0077] Each sub-verification result may be a predicted value of the verification index of the corresponding model, and whether the verification of the corresponding model is passed is determined by comparing the predicted value of the corresponding verification index and the verification condition of the corresponding verification index; schematically, the energy consumption effectiveness verification model performs energy consumption effectiveness verification on the energy consumption optimization strategy, which may refer to inputting the data corresponding to the energy consumption effectiveness verification in the energy consumption optimization strategy into the energy consumption effectiveness verification model, and obtaining the first sub-verification result of the energy consumption effectiveness verification model, that is, the predicted value of the electric energy utilization efficiency under the energy consumption optimization strategy. If the first sub-verification result meets the verification condition for the electric energy utilization efficiency, it is determined that the energy consumption effectiveness verification has passed; if the first sub-verification result does not meet the verification condition for the electric energy utilization efficiency, it is determined that the energy consumption effectiveness verification has not passed. The verification process of whether other sub-verification results indicate a pass or not can refer to the above description, which will not be repeated here.

[0078] Step 206: If the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy passes the verification, the candidate energy consumption optimization strategy is determined as the target energy consumption optimization strategy.

[0079] In a possible implementation, when the first sub-verification result, the second sub-verification result, and the third sub-verification result all indicate that the verification has passed, it is determined that the verification result indicates that the energy consumption optimization strategy has passed the verification; when there is a sub-verification result indicating that the verification has not passed among the first sub-verification result, the second sub-verification result, and the third sub-verification result, it is determined that the verification result indicates that the energy consumption optimization strategy has not passed the verification; optionally, the computer device may also set a verification ratio threshold, and when the ratio of the number of sub-verification results indicating that the verification has passed to the total number of sub-verification results is greater than the verification ratio threshold, it is determined that the verification result indicates that the energy consumption optimization strategy has passed the verification, otherwise, it is determined that the verification result indicates that the energy consumption optimization strategy has not passed the verification; the value of the verification ratio threshold can be set based on actual needs, and this application does not limit this. Schematically, taking the case where the verification result indicates that the energy consumption optimization strategy has passed the verification when the first sub-verification result, the second sub-verification result, and the third sub-verification result all indicate that the verification has passed as an example, the above-mentioned safety and effectiveness verification process for the energy consumption optimization strategy can be expressed as:

[0080] OptimalControls=Model_x*(Data_new)

[0081] (max(CurrentEnergy-Model_E(OptimalControls)) / CurrentEnergy)&(Model_C(OptimalControls)>=CurrentNeededCooling)&(MinSafeState<=Model_S(OptimalControls)<=MaxSafeState)

[0082] Among them, OptimalControls represents the target energy consumption optimization strategy, Model_x represents the x-th optimization level, Data_new represents the target operation data, CurrentEnergy represents the current energy consumption, CurrentNeededCooling represents the current cooling capacity, MinSafeState represents the minimum value of the state safety, and MaxSafeState represents the maximum value of the safety state; it should be noted that the above-mentioned setting of the verification conditions for each verification indicator is only for illustration, and other verification conditions can be set based on actual needs, and this application does not limit this.

[0083] During the hierarchical optimization process, there is a situation where the energy consumption optimization strategy of the current optimization level fails to pass the safety and effectiveness verification. In this case:

[0084] Step 207, when the target optimization level is not the highest optimization level among the optimization levels, if the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy has not passed the verification, perform the optimization calculation and safety and effectiveness verification of the next optimization level of the target optimization level.

[0085] Step 208, when the target optimization level is the highest optimization level among all optimization levels, if the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy has not passed the verification, it is determined that the HVAC equipment is not suitable for energy consumption optimization.

[0086] That is to say, when there is an optimization level higher than the target optimization level, when the target energy consumption optimization strategy cannot be obtained based on the target optimization level, the optimization process of the next optimization level is carried out. Schematically, after using the optimization model of the Nth optimization level to perform optimization calculations and safety and effectiveness verifications, if the obtained energy consumption optimization strategy passes the verification, the energy consumption optimization strategy is determined as the target energy consumption optimization strategy and output, and the hierarchical optimization process is ended; if the obtained energy consumption optimization strategy fails to pass the verification, the optimization model of the N+1th optimization level is used to continue to perform optimization calculations and safety and effectiveness verifications until the target energy consumption optimization strategy is obtained or all optimization levels are traversed, where N is a positive integer; when the target energy consumption optimization strategy is still not obtained after traversing all optimization levels, it is determined that the current HVAC equipment is not suitable for energy consumption optimization, and a corresponding prompt is output.

[0087] Since the optimization level is adjusting the parameters corresponding to each optimization index in the data set corresponding to the optimization level in the process of optimizing to achieve the purpose of reducing energy consumption, the same optimization level may output different energy consumption optimization strategies in different iterative optimizations. Schematically, the optimization index of the first optimization level is the internal control parameter of the HVAC equipment, and the internal control parameter of the HVAC equipment may include parameters such as frequency and current percentage. Different energy consumption optimization strategies can be generated based on the setting combination of different parameters. Therefore, in order to ensure the comprehensiveness of the optimization process, the computer device can perform multiple iterative optimizations at one optimization level. Based on this, when it is determined that the verification result of the candidate energy consumption optimization strategy output by the target optimization level has not passed the verification, the method also includes:

[0088] Based on the optimization model of the target optimization level, iterative optimization calculation and security and effectiveness verification are performed on the corresponding data set;

[0089] When the number of iterative optimization calculations reaches the limited number of iterations and the target energy consumption optimization strategy is not obtained, based on the level situation of the target optimization level, the optimization calculation and safety and effectiveness verification of the next optimization level of the target optimization level are performed, or it is determined that the HVAC equipment is not suitable for energy consumption optimization.

[0090] In the above process, if the target energy consumption optimization strategy is obtained when the number of iterative optimization calculations does not reach the limited number of iterations, the target energy consumption optimization strategy is output and the iterative optimization and hierarchical optimization are stopped; if the target energy consumption optimization strategy is still not obtained when the number of iterative optimization calculations reaches the limited number of iterations, it means that the current optimization level (i.e., the target optimization level) cannot perform good energy consumption optimization. When the current optimization level is not the highest optimization level, the optimization calculation and safety and effectiveness verification of the next optimization level of the current optimization level are performed. When the current optimization level is the highest optimization level, it is determined that the HVAC equipment is not suitable for energy consumption optimization. It should be noted that the value of the above-mentioned limited number of iterations can be set based on actual needs, and the setting of the limited number of iterations corresponding to different optimization levels can be the same or different, and this application does not limit this.

[0091] In summary, the energy consumption hierarchical optimization method provided in the embodiment of the present application obtains target operating data related to energy consumption optimization, performs data grading on the target operating data to obtain data sets of each optimization level corresponding to different optimization indicators, and after the starting optimization level is determined according to the operation mechanism of the equipment and the current state, the starting optimization level in each optimization level is used as the starting point for optimization, and a hierarchical optimization calculation is performed based on the optimization model of each optimization level and the data set of each optimization level to obtain a target energy consumption optimization strategy that meets the current actual needs, and the energy consumption of the HVAC equipment is optimized based on the target energy consumption optimization strategy; through the above method, the starting optimization level that is compatible with the equipment operation mechanism and the current state can be determined, and an energy consumption optimization strategy can be formulated by performing hierarchical optimization based on the starting optimization level as the starting point for hierarchical optimization, which can avoid unnecessary optimization processes and improve optimization efficiency. In addition, since the obtained target energy consumption optimization strategy is compatible with the current state and the equipment operation mechanism, the practicality of the energy consumption optimization strategy is improved.

[0092] In the process of energy consumption classification optimization, due to changes in application scenarios or application requirements, the model used in the energy consumption optimization strategy formulation may not be applicable to the current application scenarios or application requirements. At this time, the computer device can first retrain the model used in the energy consumption optimization strategy formulation, and then formulate the energy consumption optimization strategy based on the retrained model to improve the usability of the energy consumption optimization strategy. Therefore, the method also includes:

[0093] Step 209 , when it is determined that the model retraining condition is met, the historical operation data and the target operation data are combined to form a sample set.

[0094] The energy consumption grading optimization method provided in the present application involves the application of optimization models for each optimization level and safety and effectiveness verification models. In order to improve the accuracy of energy consumption grading optimization, the optimization models for each optimization level and the safety and effectiveness verification models need to be retrained under the condition that the model retraining is met; in one possible implementation method, the computer device can determine whether the conditions for model retraining are met after obtaining the target operation data related to energy consumption optimization; or, in another possible implementation method, the computer device can also determine whether model retraining is needed based on a preset judgment period, and the present application does not limit this.

[0095] Take the example of a computer device determining whether the conditions for model retraining are met after obtaining target operating data related to energy consumption optimization. In this case, after the computer device obtains the target operating data, before formulating an energy consumption optimization strategy based on the target operating data, it is necessary to first determine whether the model involved in the energy consumption optimization process needs to be retrained. If it is determined that the conditions for model retraining are met, the model retraining process is performed first, and then the energy consumption optimization strategy based on the target operating data is formulated based on the retrained model. If it is determined that the conditions for model retraining are not met, an energy consumption optimization strategy based on the target operating data can be formulated based on the current model.

[0096] The conditions for retraining the model may include at least one of the following: the current time reaches a preset training time, and the number of negative feedbacks obtained for the strategy application reaches a threshold number.

[0097] Optionally, the computer device may obtain strategy application feedback while obtaining the target operation data. The strategy application feedback may be positive feedback of the strategy application, or negative feedback of the strategy application. The computer device may count the number of times the strategy application negative feedback is sent. When it is determined that the number of times the strategy application negative feedback is obtained reaches a threshold value, it indicates that the energy consumption optimization effect of the energy consumption optimization strategy formulated by the current model is poor and cannot meet actual needs, and it is determined that the model needs to be retrained; wherein, the number of times the strategy application negative feedback is obtained may be the cumulative number of times the strategy application negative feedback is obtained within a period of time, or the number of times the strategy application negative feedback is obtained may be the number of times the strategy application negative feedback is obtained continuously.

[0098] Since the strategy formulated by the current model is no longer applicable to actual needs, if only historical operation data is used for model training, the model obtained may be more suitable for the previous state rather than the current needs. Therefore, when retraining the model, the target operation data and historical operation data are merged to form a sample set for model training.

[0099] Step 210, classify the sample data in the sample set to obtain a sample data set applied to the safety and effectiveness verification model, and a sample data set applied to the control strategy hierarchical optimization model; wherein the sample data set applied to the safety and effectiveness verification model includes a sample data subset applied to the energy consumption effectiveness verification model, a sample data subset applied to the refrigeration capacity safety verification model, and a sample data subset applied to the state safety verification; the sample data set applied to the control strategy hierarchical optimization model includes a sample data subset applied to the optimization model of each optimization level.

[0100] In the embodiment of the present application, each model is trained separately, and different models correspond to different sample data subsets. By classifying the sample data in the sample set, the sample data subset corresponding to each model is obtained; this process can be implemented as follows:

[0101] The sample data in the sample set are subjected to data classification cleaning, data classification verification, index classification fusion, index classification extraction, and index classification division to obtain a sample data set applied to the safety and effectiveness verification model, as well as a sample data set applied to the control strategy classification optimization model.

[0102] Among them, the sample data set applied to the safety and effectiveness verification includes a sample data subset applied to the energy consumption effectiveness verification model, a sample data subset applied to the refrigeration capacity safety verification model, and a sample data subset applied to the state safety verification; the data set applied to the control strategy hierarchical optimization model includes a sample data subset applied to the optimization model of each optimization level. Schematically, the sample data set applied to the safety and effectiveness verification can be expressed as {Data_E, Data_C, Data_S}, wherein Data_E represents the sample data subset applied to the energy consumption effectiveness verification model, Data_C represents the sample data subset applied to the refrigeration capacity safety verification model, and Data_S represents the sample data subset applied to the state safety verification model; the sample data set applied to the control strategy hierarchical optimization model can be expressed as {Data_01, Data_02, Data_03}, wherein Data_01 represents the sample data subset applied to the first optimization level, Data_02 represents the sample data subset applied to the second optimization level, and Data_03 represents the sample data subset applied to the third optimization level.

[0103] Among them, the relevant contents of each processing method in data classification refer to the corresponding contents in step 202, which will not be repeated here.

[0104] Step 211 , performing model training on the safety and effectiveness verification model based on the sample data set applied to the safety and effectiveness verification model, and performing model training on the control strategy hierarchical optimization model based on the sample data set applied to the control strategy hierarchical optimization.

[0105] Among them, the process of model training the safety and effectiveness verification model based on the sample data set applied to the safety and effectiveness verification model can be implemented as follows:

[0106] Performing deep learning training on the energy consumption effectiveness verification model based on a subset of sample data applied to the energy consumption effectiveness verification model to obtain a retrained energy consumption effectiveness verification model;

[0107] Performing deep learning training on the refrigeration capacity safety verification model based on a subset of sample data applied to the refrigeration capacity safety verification model to obtain a retrained refrigeration capacity safety verification model;

[0108] The state security verification model is deep-learned and trained based on a subset of sample data applied to the state security verification model to obtain a retrained state security verification model.

[0109] In one possible implementation, the computer device can use a neural network to establish a corresponding model, such as a neural network such as DNN (Deep Neural Network) or LSTM (Long Short-Term Memory); schematically, the computer device can use DNN to establish an energy consumption effectiveness verification model and a cooling capacity safety verification model, and use an LSTM network to establish a state safety verification model. It should be noted that the establishment of the above model is only schematic, and this application does not limit the type of neural network on which the model is established.

[0110] Taking the use of DNN to establish an energy consumption effectiveness verification model and a cooling capacity safety verification model, and the use of LSTM network to establish a state safety verification model as an example, the loss function used in the training process of the above three models can be expressed as:

[0111] Model_E~{min(y pue -f DNN (x Data_E )) 2}

[0112] Model_C~{min(y cooling -f DNN (x Data_C )) 2}

[0113] Model_S~{min(y state -f LSTM (x Data_S )) 2}

[0114] Among them, Model_E represents the energy consumption effectiveness verification model, Model_C represents the cooling capacity safety verification model, and Model_S represents the state safety verification model; x Data_E , x Data_C , x Data_S Respectively represent the input data of the corresponding model, y pue ,y cooling ,y stateThey respectively represent the power utilization efficiency, cooling capacity and state safety parameters of the corresponding input data. Schematically, the state safety parameters can be the inlet and return air or inlet and return water temperature, inlet and return air or inlet and return water pressure, etc. The above input data and y value are stored in the data set accordingly, so that when the model is trained, the model parameters are updated by the difference between the y value and the predicted y value obtained by the model based on the input data, thereby realizing model training and improving the accuracy of the model. When the safety and effectiveness verification model is applied, the parameters corresponding to the candidate optimization strategies determined by the control strategy hierarchical optimization model are input into each model of the safety and effectiveness verification model to obtain the corresponding sub-verification results output by each model.

[0115] The process of model training for the control strategy hierarchical optimization model based on the sample data set applied to the control strategy hierarchical optimization can be implemented as follows:

[0116] Based on the sample data subset of the first optimization level and the constraint range of the indicator parameters of the first optimization level, the optimization model of the first optimization level is trained through combined traversal, and the optimization model of the first optimization level is tested through the retrained safety and effectiveness verification model to obtain the retrained optimization model of the first optimization level;

[0117] Based on the sample data subset of the second optimization level and the constraint range of the indicator parameters of the second optimization level, the optimization model of the second optimization level is trained by Bayesian optimization, and the optimization model of the second optimization level is tested by the retrained safety and effectiveness verification model to obtain the retrained optimization model of the second optimization level;

[0118] Based on the sample data subset of the third optimization level and the indicator parameter constraint range of the third optimization level, the optimization model of the third optimization level is trained by reinforcement learning, and the optimization model of the third optimization level is tested by the retrained safety and effectiveness verification model to obtain the retrained optimization model of the third optimization level.

[0119] In an embodiment of the present application, in order to improve the retraining effect of the optimization model, the computer device may first retrain the security and effectiveness verification model, so as to test the optimization model through the retrained security and effectiveness model to obtain a retrained optimization model; wherein, for any optimization model, testing the optimization model through the retrained security and effectiveness model may refer to performing security and effectiveness verification on the parameters in the energy consumption optimization strategy obtained by the optimization model based on the corresponding sample data subset through the retrained security and effectiveness model, and when the verification result indicates that the verification is passed, it is determined that the optimization model test is passed, thereby obtaining a retrained optimization model.

[0120] In the training process of the above-mentioned optimization model, for any optimization level, a set of parameter data combinations in the sample data subset corresponding to the optimization level is input into the optimization model, and the prediction parameter data combination (i.e., the parameters in the energy consumption optimization strategy predicted by the optimization model) is obtained through the optimization model. If each parameter data in the prediction parameter data combination is within the index parameter constraint range of the optimization level, the prediction parameter data combination is verified for safety and effectiveness through the retrained safety and effectiveness model, and when the verification result indicates a pass, the above process is performed through the next set of parameter data combinations in the sample data subset until the training completion condition is met, and the retrained optimization model is obtained. Among them, the training completion condition can refer to the completion of the iterative training of the parameter data combinations of the sample data subset, or the number of times the trained optimization model passes the test of the retrained safety and effectiveness verification model is greater than the number threshold, etc.

[0121] If there are situations in which the parameter data in the prediction parameter data combination do not fall within the indicator parameter constraint range of the optimization level, or the verification result indicates failure, the model parameters of the optimization model are optimized, the process of obtaining the prediction parameter data combination is carried out again, and the subsequent process is carried out until the retrained optimization model is obtained.

[0122] In an embodiment of the present application, the constraint range of the indicator parameters of each optimization level is used to indicate the optimization space of the corresponding optimization indicator. Schematically, if the constraint range of the frequency is 50Hz to 100Hz, then the predicted frequency in the prediction parameter data combination obtained by the optimization model needs to be in the range of 50Hz to 100Hz. If it exceeds this range, it means that the optimization result of the optimization model is unreasonable and the model needs to be retrained.

[0123] Step 212, performing hierarchical optimization calculation based on the retrained control strategy hierarchical optimization model, and performing safety and effectiveness verification based on the retrained safety and effectiveness verification model.

[0124] After completing the retraining of each model, the energy consumption classification optimization strategy is formulated based on the retrained model.

[0125] By instantly updating the various models used in the energy consumption classification optimization process, the energy consumption optimization strategy obtained based on model processing can be more in line with current actual needs, thereby improving the adaptability of the energy consumption optimization strategy.

[0126] based on Figure 2 The content of the illustrated embodiment takes the process of determining whether the conditions for model retraining are met after the computer device obtains the target operation data related to energy consumption optimization, and setting three optimization levels as an example. Figure 3A flowchart of an energy consumption classification optimization method provided by an exemplary embodiment of the present application is shown. The method can be executed by a computer device, and the computer device can be implemented as a management device for HVAC equipment, such as Figure 3 As shown, the energy consumption classification optimization method may include:

[0127] S301, obtaining target operation data.

[0128] S302, determine whether the conditions for model retraining are met, if not, execute S303, otherwise, execute S307.

[0129] The conditions for retraining the model may include at least one of the following: the current time reaches the preset training time, and the number of negative feedbacks obtained for the strategy application reaches a threshold number; wherein, when the conditions for retraining the model include a judgment on the number of negative feedbacks, the computer device may obtain strategy application feedback when obtaining target operation data.

[0130] S303, determining whether the target operation data passes the optimization pre-detection, if so, executing S304, otherwise, determining that the HVAC equipment is not suitable for energy consumption optimization.

[0131] S304, classifying the target operation data.

[0132] This process can refer to Figure 2 The relevant contents in the illustrated embodiment will not be repeated here.

[0133] S305, based on the equipment operation mechanism, that is, the current state indicated by the target operation data, determine the starting optimization level and perform hierarchical optimization calculation.

[0134] This process can refer to Figure 2 The relevant contents in the illustrated embodiment will not be repeated here.

[0135] S306: Output the target energy consumption optimization strategy.

[0136] S307, merging the historical operation data and the target operation data as a sample set.

[0137] S308, performing data classification on the sample data in the sample set to obtain a sample data set applied to a safety and effectiveness verification model, and a sample data set applied to a control strategy classification optimization model.

[0138] S309, performing model training on the safety and effectiveness verification model based on the sample data set applied to the safety and effectiveness verification model.

[0139] S310: Performing model training on a control strategy hierarchical optimization model based on a sample data set applied to the control strategy hierarchical optimization.

[0140] After obtaining the retrained safety and effectiveness verification model and the control strategy hierarchical optimization model, S305 is executed to obtain a target energy consumption optimization strategy for the target operation data.

[0141] The implementation method of the above process can refer to Figure 2 The relevant contents of the embodiment will not be repeated here.

[0142] Figure 4 A schematic diagram of an energy consumption classification optimization system provided by an exemplary embodiment of the present application is shown. The system is used to perform the following steps: Figure 1 , Figure 2 or Figure 3 All or part of the steps of the embodiment shown in the figure can be used to obtain the target energy consumption optimization strategy, such as Figure 4 As shown, the energy consumption classification optimization system may include a data classification service module 410, a data classification processing module 420, a safety and effectiveness verification module 430 and a control strategy classification optimization module 440. The data classification processing module 420, the safety and effectiveness verification module 430 and the control strategy classification optimization module 440 constitute an energy consumption classification intelligent optimization module.

[0143] Among them, the data classification service module 410 is used to obtain target operation data. The data classification service module can serve as an intermediary between the real data center and the energy consumption classification intelligent optimization module that performs energy consumption classification optimization. On the one hand, the data classification service module can collect various indoor and outdoor environmental data, equipment operation status data, equipment control setting parameters, control feedback data, equipment combination operation mode, equipment operation quantity, IT center load and energy consumption data, refrigeration equipment and power supply energy consumption data, etc. from the real data center; considering the need for subsequent classification processing, the data classification module can also perform classification storage, classification forwarding and constraint classification configuration on the acquired target operation data after obtaining the target operation data; on the other hand, the data classification service module is also responsible for sending the target energy consumption optimization strategy obtained by the energy consumption classification intelligent optimization module to the real data center.

[0144] The data classification processing module 420 is used to perform data classification processing on the input operation data, including data classification cleaning, data classification verification, indicator classification fusion, indicator classification extraction, and indicator classification division, so as to obtain a data set applied to the safety and effectiveness verification model and a data set applied to the control strategy classification optimization model, and send the corresponding data sets to the safety and effectiveness verification module 430 and the control strategy classification optimization module 440 respectively.

[0145] The safety and effectiveness verification module 430 can be used to train the safety and effectiveness verification model and to verify the safety and effectiveness of the energy consumption optimization strategy through the safety and effectiveness verification model.

[0146] The control strategy hierarchical optimization module 440 can be used to train the control strategy hierarchical optimization model, determine the starting optimization level, and perform hierarchical optimization calculations based on the optimization models of each optimization level to obtain a target energy consumption optimization strategy.

[0147] Figure 5 The structural block diagram of a computer device 500 shown in an exemplary embodiment of the present application is shown. The computer device can be implemented as a management device for HVAC equipment in the above-mentioned solution of the present application. The computer device 500 includes a central processing unit (CPU) 501, a system memory 504 including a random access memory (RAM) 502 and a read-only memory (ROM) 503, and a system bus 505 connecting the system memory 504 and the central processing unit 501. The computer device 500 also includes a large-capacity storage device 506 for storing an operating system 509, an application program 510 and other program modules 511.

[0148] Without loss of generality, the computer-readable medium may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Computer storage media include RAM, ROM, Erasable Programmable Read Only Memory (EPROM), Electronically-Erasable Programmable Read-Only Memory (EEPROM) flash memory or other solid-state storage technology, CD-ROM, Digital Versatile Disc (DVD) or other optical storage, cassettes, tapes, disk storage or other magnetic storage devices. Of course, those skilled in the art will appreciate that the computer storage medium is not limited to the above. The above-mentioned system memory 504 and mass storage device 506 can be collectively referred to as memory.

[0149] According to various embodiments of the present application, the computer device 500 can also be connected to a remote computer on the network through a network such as the Internet. That is, the computer device 500 can be connected to the network 508 through the network interface unit 507 connected to the system bus 505, or the network interface unit 507 can be used to connect to other types of networks or remote computer systems (not shown).

[0150] The memory also includes at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is stored in the memory. The central processing unit 501 implements all or part of the steps in the energy consumption classification optimization method shown in the above-mentioned embodiments by executing the at least one instruction, at least one program, code set or instruction set.

[0151] In an exemplary embodiment, a computer-readable storage medium is also provided, in which at least one computer program is stored, and the computer program is loaded and executed by a processor to implement all or part of the steps in the application interaction method during the above call process. For example, the computer-readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, etc.

[0152] In an exemplary embodiment, a computer program product is also provided, the computer program product comprising at least one computer program, the computer program being loaded and executed by a processor Figure 1 , Figure 2 ,or Figure 3 All or part of the steps of the application interaction method during a call as shown in any embodiment.

[0153] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.

[0154] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for optimizing energy consumption by grading, characterized in that: The method comprises: Obtain target operating data related to energy consumption optimization; The target operation data is classified to obtain data sets corresponding to various optimization levels; data sets at different optimization levels correspond to different optimization indicators; Determining a starting optimization level based on the equipment operation mechanism and the current state indicated by the target operation data; Taking the initial optimization level as the starting point for optimization, hierarchical optimization calculations are performed based on the optimization models of each optimization level and the data sets corresponding to each optimization level to obtain a target energy consumption optimization strategy, so as to optimize the energy consumption of HVAC equipment based on the target energy consumption optimization strategy.

2. The method according to claim 1, characterized in that The method takes the initial optimization level as the optimization starting point, performs hierarchical optimization calculation based on the optimization models of each optimization level and the data sets corresponding to each optimization level, and obtains the target energy consumption optimization strategy, including: When entering the target optimization level, the optimization model of the target optimization level is used to perform optimization calculation on the corresponding data set to obtain a candidate energy consumption optimization strategy; the target optimization level is any one of the optimization levels that uses the initial optimization level as the optimization starting point for optimization; Performing safety and effectiveness verification on the candidate energy consumption optimization strategy through a safety and effectiveness verification model to obtain a verification result of the candidate energy consumption optimization strategy; If the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy passes the verification, the candidate energy consumption optimization strategy is determined as the target energy consumption optimization strategy.

3. The method according to claim 2, characterized in that The method further comprises: In the case where the target optimization level is a non-highest optimization level among the optimization levels, if the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy has not passed the verification, performing optimization calculation and safety and effectiveness verification of the next optimization level of the target optimization level; In the case where the target optimization level is the highest optimization level among the optimization levels, if the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy has not passed the verification, it is determined that the HVAC equipment is not suitable for energy consumption optimization.

4. The method according to claim 3, characterized in that In a case where the verification result of the candidate energy consumption optimization strategy indicates that the candidate energy consumption optimization strategy fails the verification, the method further includes: Based on the optimization model of the target optimization level, iterative optimization calculation and security and effectiveness verification are performed on the corresponding data set; When the number of iterative optimization calculations reaches the specified number of iterations and the target energy consumption optimization strategy is not obtained, based on the level situation of the target optimization level, the optimization calculation and safety and effectiveness verification of the next optimization level of the target optimization level are performed, or it is determined that the HVAC equipment is not suitable for energy consumption optimization.

5. The method according to claim 1, characterized in that: The target operation data includes indoor and outdoor temperature and humidity and information technology IT load changes; The determining of the initial optimization level based on the equipment operation mechanism and the current state indicated by the target operation data includes: When the outdoor temperature is within the preset outdoor temperature range, the indoor and outdoor temperature and humidity change is periodic, and the IT load change is periodic, the starting optimization level is determined to be the first optimization level; the first optimization level indicates to optimize the internal control parameters of the HVAC equipment; When the outdoor temperature is within the outdoor temperature range, the indoor and outdoor temperature and humidity changes indicate that there is a sudden change in the indoor and outdoor temperature and humidity, or the IT load changes indicate that there is a sudden change in the IT load, determining that the starting optimization level is the second optimization level; the second optimization level indicates that the operating quantity and internal control parameters of the HVAC equipment are optimized; When the outdoor temperature is within a critical temperature range between different outdoor temperature ranges, the starting optimization level is determined to be a third optimization level; the third optimization level indicates optimizing the operation mode, operation quantity and internal control parameters of the HVAC equipment.

6. The method according to claim 1, characterized in that The step of grading the target operation data to obtain data sets corresponding to each optimization level includes: The target operation data is subjected to data classification cleaning, data classification verification, index classification fusion, index classification extraction, and index classification division to obtain a data set applied to a safety and effectiveness verification model; the control strategy classification optimization model includes optimization models for each optimization level; the data set applied to the control strategy classification optimization model includes a data subset of the optimization model applied to each optimization level; Among them, the data hierarchical verification is used to verify the correctness of each sample data; the indicator hierarchical fusion is used to perform data statistics on related indicators and generate new indicators; the indicator hierarchical extraction is used to extract the indicator sets required by different models to form corresponding data subsets; the indicator hierarchical division is used to classify each indicator into types.

7. The method according to claim 2, characterized in that: The safety and effectiveness verification model includes an energy consumption effectiveness verification model, a refrigeration capacity safety verification model and a state safety verification model; The energy consumption optimization strategy obtained by the optimization model is verified for safety and effectiveness through the safety and effectiveness verification model, including: Performing energy efficiency verification on the energy optimization strategy through the energy efficiency verification model to obtain a first sub-verification result; The cooling capacity safety verification model is used to verify the cooling capacity safety of the energy consumption optimization strategy, and a second sub-verification result is obtained; The energy consumption optimization strategy is verified for state safety through the state safety verification model to obtain the third sub-verification result; A verification result is generated based on the first sub-verification result, the second sub-verification result, and the third sub-verification result; the verification result indicates whether the energy consumption optimization strategy has passed the verification or has failed the verification.

8. The method according to claim 1, characterized in that: The method further comprises: When it is determined that the model retraining condition is met, the historical operation data is combined with the target operation data to form a sample set; The sample data in the sample set are classified to obtain a sample data set applied to a safety and effectiveness verification model, and a sample data set applied to a control strategy hierarchical optimization model; wherein the sample data set applied to the safety and effectiveness verification model includes a sample data subset applied to an energy consumption effectiveness verification model, a sample data subset applied to a refrigeration capacity safety verification model, and a sample data subset applied to a state safety verification; the sample data set applied to a control strategy hierarchical optimization model includes a sample data subset applied to optimization models at each optimization level; Performing model training on the safety and effectiveness verification model based on the sample data set applied to the safety and effectiveness verification model, and performing model training on the control strategy hierarchical optimization model based on the sample data set applied to the control strategy hierarchical optimization; A hierarchical optimization calculation is performed based on the retrained control strategy hierarchical optimization model, and a safety and effectiveness verification is performed based on the retrained safety and effectiveness verification model.

9. The method according to claim 8, characterized in that The safety and effectiveness verification model is trained based on the sample data set applied to the safety and effectiveness verification model, including: Performing deep learning training on the energy consumption effectiveness verification model based on a subset of sample data applied to the energy consumption effectiveness verification model to obtain a retrained energy consumption effectiveness verification model; Performing deep learning training on the refrigeration capacity safety verification model based on a subset of sample data applied to the refrigeration capacity safety verification model to obtain a retrained refrigeration capacity safety verification model; The state security verification model is deep-learned and trained based on a subset of sample data applied to the state security verification model to obtain a retrained state security verification model.

10. The method according to claim 8, characterized in that The method of training the control strategy hierarchical optimization model based on the sample data set applied to the control strategy hierarchical optimization includes: Based on the sample data subset of the first optimization level and the constraint range of the indicator parameters of the first optimization level, the optimization model of the first optimization level is trained through combined traversal, and the optimization model of the first optimization level is tested through the retrained safety and effectiveness verification model to obtain the retrained optimization model of the first optimization level; Based on the sample data subset of the second optimization level and the constraint range of the indicator parameters of the second optimization level, the optimization model of the second optimization level is trained by Bayesian optimization, and the optimization model of the second optimization level is tested by the retrained safety and effectiveness verification model to obtain the retrained optimization model of the second optimization level; Based on the sample data subset of the third optimization level and the constraint range of the indicator parameters of the third optimization level, the optimization model of the third optimization level is trained by reinforcement learning, and the optimization model of the third optimization level is tested by the retrained safety and validity verification model to obtain the retrained optimization model of the third optimization level.

11. The method according to claim 8, characterized in that The condition for retraining the model includes at least one of the following: the current time reaches the preset training time, and the number of negative feedbacks obtained for the strategy application reaches a threshold number.

12. The method according to claim 1, characterized in that Before classifying the target operation data to obtain data sets corresponding to each optimization level, the method further includes: Performing optimization pre-detection on the target operation data through optimization pre-detection indicators to obtain optimization pre-detection results; the optimization pre-detection indicators include at least one of the following: detecting whether the indicators required by each model are missing and whether the business relationship between the indicators is established; The step of grading the target operation data to obtain data sets corresponding to each optimization level includes: When the optimization pre-detection result indicates that the target operation data has passed the pre-detection, the target operation data is classified to obtain data sets corresponding to each optimization level.

13. The method according to claim 12, characterized in that The method further comprises: When the optimization pre-detection result indicates that the target operation data detection fails, it is determined that the HVAC equipment is not suitable for energy consumption optimization.

14. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the energy consumption hierarchical optimization method according to any one of claims 1 to 13.

15. A computer-readable storage medium, characterized in that: At least one computer program is stored in the computer-readable storage medium, and the computer program is loaded and executed by the processor to implement the energy consumption classification optimization method according to any one of claims 1 to 13.

Images