Building energy efficiency optimization system and method

Abstract

The invention relates to the field of artificial intelligence and energy efficiency control, and discloses a building energy efficiency optimization system and method, and the method comprises the steps: obtaining historical load data, real-time operation state data and target time meteorological data of a heating and ventilation system; inputting the historical load data, the real-time operation state data and the meteorological data of the target time into a dual-core optimization model constructed based on a Transformer-genetic algorithm and an expert rule to obtain an optimal control strategy; the power consumption corresponding to the optimal control strategy is lowest; a heating and ventilation system is controlled to execute an optimal control strategy, a dual-core-driven decision system is constructed, deterministic control logic based on expert knowledge is organically combined with a deep learning optimization algorithm based on mass data, a control center with complementary advantages and system operation is formed, meanwhile, the energy efficiency optimization effect corresponding to the control strategy is improved, and the system performance is improved. And low-energy-efficiency, high-comfort and high-safety control strategy output is realized.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of artificial intelligence and energy efficiency control, in particular to a building energy efficiency optimization system and method. BACKGROUND

[0002] The energy consumption system of a public building is complex, and is generally divided into air conditioning, lighting sockets, elevators, power, special equipment and other energy consumption systems. Among them, the heating, ventilation and air conditioning system (HVAC) is the largest energy consumption system in public buildings, and its power consumption accounts for as high as 40%-60%. At the same time, because of the problems such as system design redundancy, complex equipment operation, lack of control strategy, and insufficient professional operation personnel, the heating, ventilation and air conditioning system has become the system with the largest energy saving potential in public buildings, and the mainstream energy management system in the current industry also focuses on the power consumption of the heating, ventilation and air conditioning system.

[0003] In the related art, the energy management system mostly stays in the "visible but not intelligent" stage, and its core function is limited to data board display and standardized report generation. However, for the key needs of power consumption anomaly hiding and optimization strategy generation, the system often only provides "excessive energy consumption" alarms. SUMMARY

[0004] In view of the above defects or deficiencies in the prior art, it is desirable to provide a building energy efficiency optimization system and method to construct a "dual-core driven" decision system, organically combine the deterministic control logic based on expert knowledge with the deep learning optimization algorithm based on massive data, form complementary advantages, and form the control center of the system operation, while improving the energy efficiency optimization effect of the corresponding control strategy, and realizing the output of the control strategy with low energy efficiency, high comfort and high safety.

[0005] In a first aspect, an embodiment of the present application provides a building energy efficiency optimization method, comprising: obtaining historical load data, real-time operation state data and meteorological data of a target time of a heating system; inputting the historical load data, the real-time operation state data and meteorological data of the target time into a dual-core optimization model based on a Transformer-genetic algorithm and an expert rule to obtain an optimal control strategy; the optimal control strategy corresponds to the lowest power consumption; wherein the dual-core optimization model comprises a Transformer-genetic algorithm sub-model and an expert rule sub-model; the Transformer-genetic algorithm sub-model is used to predict a predicted load at the target time based on the historical load data, the real-time operation state data and the meteorological data of the target time, and to optimize a control strategy based on the predicted load by using a genetic algorithm, and the expert rule sub-model is used to correct the control strategy in the optimization process based on a preset expert rule during the process in which the Transformer-genetic algorithm sub-model optimizes the control strategy based on the predicted load by using the genetic algorithm. controlling the heating and ventilation system to execute the optimal control strategy.

[0006] In some embodiments, the optimization of the control strategy based on the predicted load by using the genetic algorithm comprises: randomly generating a first preset number of initial control strategies and obtaining an adaptive score corresponding to each initial control strategy; randomly extracting a second preset number of the initial control strategies, selecting the initial control strategy with the highest adaptive score as a genetic parent control strategy, and repeating the above process until a third preset number of genetic parent control strategies are obtained; randomly extracting any two genetic parent control strategies for cross breeding to obtain a genetic child control strategy corresponding to the genetic parent control strategies, and repeating the above process until a fourth preset number of genetic child control strategies are obtained; obtaining a child power consumption value corresponding to each genetic child control strategy in the current round, and performing cross breeding in the next round until the difference between the child power consumption value corresponding to the current round and the child power consumption value corresponding to the previous round is less than a preset error.

[0007] In some embodiments, the obtaining of the adaptive score corresponding to each initial control strategy comprises: obtaining the power consumption of a chiller, a chilled water pump, a cooling water pump and a cooling tower under the initial control strategy respectively, and calculating the total power consumption of the initial control strategy; calculating a first rule penalty factor and a second rule penalty factor corresponding to the initial control strategy according to the preset expert rule, and determining an expert rule penalty term according to the first rule penalty factor and the second rule penalty factor; determining the adaptive score according to the total power consumption and the expert rule penalty term.

[0008] In some embodiments, the strategy modification of the control strategy in the optimization process based on the preset expert rules comprises: In the process of the random extraction of any two of the genetic parent control strategies for cross breeding, the safe-oriented cross breeding of the two genetic parent control strategies is based on the preset expert rules.

[0009] In some embodiments, the safe-oriented cross breeding of the two genetic parent control strategies comprises: Any one of the genetic parent control strategies is selected from the two genetic parent control strategies, and the safety evaluation of each gene segment of the genetic parent control strategy is based on the preset expert rules. According to the safety evaluation value, a cross point is selected. According to the cross point, the gene segments of the two genetic parent control strategies are exchanged to obtain two genetic child control strategies.

[0010] In some embodiments, the selection of the cross point according to the safety evaluation value comprises: According to the preset expert rules, the dependency relationship between the gene segments is obtained. According to the dependency relationship, a gene segment group with a dependency relationship and the gene position of each gene segment in the gene segment group are determined. According to the gene position of each gene segment in the gene segment group, the cross point is selected.

[0011] In a second aspect, the embodiments of the present application provide a building energy efficiency optimization system, comprising: An acquisition module is configured to acquire historical load data, real-time operation state data and meteorological data of a target time of a heating and ventilation system. The dual-core driving module is configured to input the historical load data, the real-time operation state data and the meteorological data of the target time into a dual-core optimization model based on a Transformer-genetic algorithm and an expert rule to obtain an optimal control strategy; the optimal control strategy corresponds to the lowest power consumption; wherein the dual-core optimization model comprises a Transformer-genetic algorithm sub-model and an expert rule sub-model; the Transformer-genetic algorithm sub-model is configured to predict a predicted load of the target time based on the historical load data, the real-time operation state data and the meteorological data of the target time, and optimize a control strategy based on the predicted load by using a genetic algorithm, and the expert rule sub-model is configured to correct the control strategy in the optimization process based on a preset expert rule during the process in which the Transformer-genetic algorithm sub-model optimizes the control strategy based on the predicted load by using the genetic algorithm. The control module is configured to control the heating and ventilation system to execute the optimal control strategy.

[0012] In a third aspect, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method described in the embodiments of the present application when executing the program.

[0013] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the program is executable on a processor to implement the method described in the embodiments of the present application.

[0014] In a fifth aspect, a computer program product is provided, which includes a computer program, and the computer program is executable on a processor to implement the method described in the embodiments of the present application.

[0015] The building energy efficiency optimization system and method provided by the embodiments of the present application can obtain historical load data, real-time operation state data and meteorological data of a target time, input the historical load data, the real-time operation state data and the meteorological data of the target time into a dual-core optimization model based on a Transformer-genetic algorithm and an expert rule to obtain an optimal control strategy, control the heating and ventilation system to execute the optimal control strategy, and realize the output of a control strategy with low energy efficiency, high comfort and high safety.

[0016] Additional aspects and advantages of the application will be described in the description that follows, and will become apparent from the description, or will be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0017] Other features, objects, and advantages of the application will become more apparent from the following detailed description when read in conjunction with the accompanying drawings: Figure 1 An implementation environment architecture diagram of a building energy efficiency optimization method provided by an embodiment of the application is shown; Figure 2 A flowchart of a building energy efficiency optimization method provided by an embodiment of the application is shown; Figure 3 A flowchart of a building energy efficiency optimization method provided by another embodiment of the application is shown; Figure 4 A structural diagram of a building energy efficiency optimization system provided by an embodiment of the application is shown; Figure 5 A structural diagram of a computer system of an electronic device or server suitable for implementing an embodiment of the application is shown. DETAILED DESCRIPTION

[0018] The application will be further described below in conjunction with the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related application, and not to limit the application. In addition, it should be noted that only the parts related to the application are shown in the drawings for ease of description.

[0019] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0020] The total building operation power consumption in China is 11.9 tce, accounting for 22% of the total energy consumption in China. The total building operation carbon emission in China is 23.1 billion tco2, accounting for 21.7% of the total energy-related carbon emission in China. Among them, the operation power consumption of public buildings is 4.9 billion tce (accounting for 41%), and the operation carbon emission of public buildings is 9.4 billion tco2 (accounting for 41%), which is the main source of building operation power consumption and carbon emission. Under the drive of the "double carbon" target, low-carbon operation of public buildings has become a key battlefield for reducing the overall carbon emission of society.

[0021] Currently, the mainstream energy management systems in the industry include Honeywell's Building Energy Management Suite (Honeywell BeMS), Siemens' SSE-EMS, and Schneider Electric's EMS (Energy Management System). Honeywell's intelligent building energy management system, as an integrated solution, combines energy management with energy-saving measures. It monitors the status of high-power equipment such as HVAC, lighting, water supply and drainage, and power distribution within buildings, provides equipment protection and operation management, and evaluates and analyzes power consumption data. The system features energy flow analysis and display, multi-dimensional indicator statistical comparison, power consumption alarms, and demand control. The SSEP-EMS system records all power consumption data in the factory through various field instruments or sensors, such as smart meters and water meters, and uses intuitive load curves to quickly and accurately display energy consumption. The system also features energy flow analysis and display, peak-valley-flat management, and power consumption alarms. Schneider EMS is a software platform specifically designed for energy efficiency management. It helps users ensure safer, more reliable, and more efficient electricity use by calculating, modeling, predicting, and tracking performance indicators for all energy sources (including electricity, water, and gas) through energy visualization and analytics tools. Schneider EMS assists users in energy audits, indicator analysis, and cost allocation to further evaluate, advance, and validate energy-saving effects, and promptly identify abnormal energy usage throughout the process.

[0022] It is evident that current energy management systems remain at the "visible but not intelligent" stage. When an alarm for "excessive energy consumption" is triggered, professional technicians are required to analyze the data in order to resolve the problem.

[0023] Based on this, this application eliminates a building energy efficiency optimization system and method, and constructs a "dual-core driven" decision-making system. It organically combines deterministic control logic based on expert knowledge with deep learning optimization algorithms based on massive data to form a complementary advantage and a control center for system operation. At the same time, it improves the energy efficiency optimization effect corresponding to the control strategy and achieves the output of control strategies with low energy efficiency, high comfort, and high safety.

[0024] For the specific implementation environment of the building energy efficiency optimization method proposed in this application, please refer to [link / reference needed]. Figure 1 . Figure 1 The diagram illustrates the implementation environment architecture of the building energy efficiency optimization method provided in this application embodiment.

[0025] like Figure 1 As shown, the implementation environment architecture includes: an energy efficiency management system 101 and a building automation system 102.

[0026] The building automation system (BAS) 102 is configured to collect the running states of devices in the building heating system and sensor parameters (for example, water pipe temperature, valve opening, fan frequency, etc.) in real time and send them to the energy management system (EMS) 101. The energy management system 101 is configured to execute the building energy efficiency optimization method provided in the embodiments of the present application, determine the optimal control strategy with the optimal energy efficiency, and send the optimal control strategy to the building automation system 102 for accurate execution by the building automation system 102.

[0027] The energy management system 101 can be a physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, and big data and artificial intelligence platform.

[0028] The energy management system 101 and the building automation system 102 are directly or indirectly connected through wired or wireless communication. Optionally, the wireless network or wired network uses standard communication technology and / or protocol. The network is usually the Internet, but can also be any network, including but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or any combination of virtual private networks.

[0029] The building energy efficiency optimization method provided in the present application can be implemented by a building energy efficiency optimization device, which can be installed on a terminal device or a server.

[0030] In order to further illustrate the technical solutions provided in the embodiments of the present application, the following will be described in detail in conjunction with the drawings and specific embodiments. Although the embodiments provided in the present application include the following method operation instruction steps, more or fewer operation instruction steps can be included in the method based on conventional or non-creative labor. The execution order of the steps is not limited to the execution order provided in the embodiments of the present application in the logical sense. The method can be executed in sequence or in parallel during actual processing or device execution.

[0031] It should be noted that the data obtained or used in the embodiments of the present application needs to be agreed by the user, and the relevant data can be obtained only after the user's authorization and permission, and the data obtained or used complies with the relevant legal regulations.

[0032] Reference is made to Figure 2 , Figure 2 A flowchart of a building energy efficiency optimization method provided by an embodiment of the present application is shown. As shown in Figure 2 , the method comprises: Step 201, obtaining historical load data, real-time running state data and meteorological data of a target time of a heating and ventilation system.

[0033] It should be noted that, since the prior art usually focuses on “visible” information, it usually only focuses on the time series data of total power consumption, and lacks insight into the driving factors behind energy consumption behavior. Based on this, the present application proposes to use multi-dimensional data to predict the regularity of the load at the target time.

[0034] Among them, the historical load data includes historical equipment running data, historical environmental parameter data and historical power consumption time series data. The historical equipment running data includes the start-stop state, running power, frequency, time length, etc. of the heating and ventilation host, water pump, fan, lighting circuit and other equipment; the historical environmental parameter data includes indoor and outdoor temperature, humidity, light intensity, CO2 concentration, crowd density, etc.; the power consumption time series data includes the power consumption history data of electricity, water, gas, etc. of each region, each classification and each item.

[0035] Step 202, inputting the historical load data, real-time running state data and meteorological data of the target time into a dual-core optimization model based on a Transformer-genetic algorithm and expert rules to obtain an optimal control strategy; wherein the dual-core optimization model comprises a Transformer-genetic algorithm sub-model and an expert rule sub-model; the Transformer-genetic algorithm sub-model is used to predict the predicted load of the target time based on the historical load data, real-time running state data and meteorological data of the target time, and to optimize the control strategy based on the predicted load using a genetic algorithm, and the expert rule sub-model is used to correct the control strategy in the optimization process based on expert rules during the process of the Transformer-genetic algorithm sub-model optimizing the control strategy based on the predicted load using the genetic algorithm.

[0036] Step 203, controlling the heating and ventilation system to execute the optimal control strategy.

[0037] Specifically, the present application constructs a “dual-core driven” decision system, organically combines the deterministic control logic based on expert knowledge and the deep learning optimization algorithm of massive historical data, and forms a control center with complementary advantages and collaborative operation.

[0038] The first driving core is an expert rule sub-model, which is used to provide expert rules and basic strategies. The driving is used to provide basic safety for the system and control strategy according to the preset expert rules. Specifically, the expert rules and basic strategies provided by the expert rule sub-model are the “safety cornerstone” and “experience base” of the entire system. The core task is to define a high safety and high reliability operation boundary for the system, and to provide industry-recognized and effective basic energy-saving strategies. Using the preset expert rules to optimize the control strategy in the optimization process can effectively ensure that the control strategy generated in the optimization process will not trigger dangerous operations, solve the cold start problem of the system when there is a lack of sufficient training data, and provide clear and traceable explanations for the basic operation logic of the system.

[0039] The second driving core is a Transformer-genetic algorithm sub-model, which is used to predict the predicted load of the target time using the Transformer model, and to optimize the control strategy based on the predicted load using the genetic algorithm. In building energy management systems, short-term prediction of cold load is the basis of active optimization control. Accurate prediction of the future trend of cold load change at future time helps to realize dynamic optimization control of cold machine group start-stop strategy, cold storage system charging and discharging, primary / secondary pump flow regulation and other links, thereby effectively improving the energy efficiency and response flexibility of the system. Especially when the load presents obvious nonlinear, periodic and multi-factor driven (such as weather, time period, operation mode) characteristics, it is difficult to obtain ideal prediction accuracy by relying on traditional rules or linear models.

[0040] The cold load prediction task has significant time sequence, periodicity and multi-source driving characteristics. The change is not only affected by weather factors (such as outdoor temperature and humidity), but also closely related to the user behavior pattern, operation strategy and historical load state of the building itself. In order to accurately depict the influence of these complex factors on future load, the prediction model needs to have strong sequence modeling capability and multi-variable feature fusion capability. The Transformer model is originally applied to the field of natural language processing, and its core structure is based on the self-attention mechanism (Self-Attention), which can model the global dependency relationship between any positions in the sequence without relying on the time sequence. In recent years, Transformer has gradually shown superior performance in time series prediction tasks and has been widely used in the fields of power, transportation, meteorology and the like. The present application introduces it into the building cold load prediction task. The Transformer model takes the self-attention mechanism as the core and can establish a direct information channel between any two time positions in the sequence, thereby effectively capturing the long-time span dependency relationship. This feature enables it to model the periodic disturbance and response to sudden events without relying on explicit time sequence, and is particularly suitable for modeling the nonlinear process of cold load which is significantly affected by sunlight, time period, behavior and the like. In addition, the Transformer allows multiple feature channels to be input in structure, and the multi-head attention mechanism in the model can simultaneously learn the interaction relationship between features from different dimensions, having good high-dimensional feature fusion capability. Combined with appropriate embedding coding mode, the Transformer can input meteorological data, historical load, operation state, time stamp and the like information together and complete feature expression and dependency modeling in a unified structure, effectively avoiding the problem of insufficient integration of multi-source information in traditional models.

[0041] And the genetic algorithm is an optimization algorithm based on the principles of natural selection and genetics, which has the advantages of strong global search capability, wide application range, strong parallelism, high robustness, and easy extension. In the process of optimizing the control strategy based on the predicted load by using the genetic algorithm, the control strategy in the optimization process is modified based on the preset expert rules, realizing genetic optimization based on a safety-oriented operator, greatly improving the safety of the control strategy optimization process and improving the control strategy optimization efficiency.

[0042] Therefore, the building energy efficiency optimization method provided by the embodiments of the present application inputs the historical load data, real-time operation state data and meteorological data of the target time into the dual-core optimization model constructed based on the Transformer-genetic algorithm and expert rules to obtain the optimal control strategy, and the optimal control strategy corresponds to the lowest power consumption; and controls the HVAC system to execute the optimal control strategy, realizing the output of the control strategy with low energy efficiency, high comfort and high safety.

[0043] In some embodiments, as shown in Figure 3 illustrated, a genetic algorithm is utilized to optimize the control strategy based on the predicted load, including: Step 301, a first preset number of initial control strategies are randomly generated, and the fitness score corresponding to each initial control strategy is obtained.

[0044] It should be noted that the initial control strategy is an initial solution for genetic optimization, and each initial control strategy includes a complete control strategy and a control scheme for all control variables at all time periods before the target time.

[0045] For example, the control strategy X can be expressed as follows:

[0046] wherein, represents the state of the i-th device 1, 2…k, represents the state of the j-th parameter j 1, 2…m.

[0047] It should be further noted that the first preset number is set according to the actual situation of the building HVAC system, so that the first preset number of initial control strategies can cover the feasible solution space of the HVAC system control strategy, and provide a space foundation for the subsequent optimization process.

[0048] Further, the present application also calculates the fitness score of each initial control strategy to calculate the power consumption and violation of constraints of each initial control strategy, and gives a comprehensive score. The higher the score, the better the strategy, i.e. low and meeting the constraints.

[0049] In some embodiments, the fitness score corresponding to each initial control strategy is obtained, including: the power consumption of the chiller unit, the chilled water pump, the cooling water pump and the cooling tower under the initial control strategy is obtained respectively, and the total power consumption of the initial control strategy is calculated; the first rule penalty factor and the second rule penalty factor corresponding to the initial control strategy are calculated according to the preset expert rule, and the expert rule penalty term is determined according to the first rule penalty factor and the second rule penalty factor; the fitness score is determined according to the total power consumption and the expert rule penalty term.

[0050] wherein, for the chiller unit, the power consumption can be expressed as:

[0051] wherein, is the power consumption of the o-th chiller unit, is the rated power of the o-th chiller unit, Let be the partial load rate of the o-th chiller unit at time t. The temperature of the cooling water inlet for the o-th chiller unit is .

[0052] For chilled water pumps or cooling water pumps, their power consumption can be expressed as:

[0053] in, The power of the water pump refers to the pump's capacity; the water pumps include chilled water pumps and cooling water pumps. The density of water, It is the acceleration due to gravity. and These are the instantaneous flow rate and head of the p-th pump at time t, respectively. Let be the overall hydraulic efficiency of the p-th pump under the initial control strategy.

[0054] For a cooling tower, its power consumption can be expressed as:

[0055] in, For the qth cooling tower at the rated frequency The power below, Let be the operating frequency of the q-th cooling tower at time t under the initial control strategy.

[0056] Therefore, the total power consumption can be expressed as:

[0057] in, Let be the total power consumption at time t under the initial control strategy. This represents the start-up and shutdown status of the o-th chiller unit at time t. This represents the total number of chiller units. For chilled water pumps, This refers to the total number of chilled water pumps. For cooling water pumps, This represents the total number of cooling water pumps. This represents the total number of cooling towers.

[0058] Preferably, time t is the target time, i.e. This represents the total power consumption at the target time under the initial control strategy.

[0059] Furthermore, the first rule is a pre-defined expert rule that must never be violated, while the second rule is a pre-defined expert rule that can be disregarded. For example, the first rule could be that if the water pump is turned off, the valve will close, or the temperature limit for chilled water, etc., while the second rule could be the building's internal temperature, etc.

[0060] For example, an expert rule penalty term can be represented as:

[0061] wherein, is a penalty term of the expert rule, is a penalty weight of the a-th first rule, is a first rule constraint function of the a-th first rule, is a total number of the first rules, is a penalty weight of the b-th second rule, is a constraint function of the b-th second rule, is a target value corresponding to the b-th second rule, is an initial control strategy, wherein, > .

[0062] Thus, the fitness score can be expressed as:

[0063] wherein, is the fitness score, is the total power consumption, is a dynamic penalty coefficient, is a penalty term of the expert rule, is a smoothing constant to prevent division by zero.

[0064] In some preferred embodiments, the dynamic penalty coefficient can be dynamically adjusted, and an exemplary expression can be related to the outdoor temperature, and can be expressed as:

[0065] wherein, is the dynamic penalty coefficient, is a basic penalty coefficient, is a temperature sensitivity coefficient, is an outdoor temperature deviation (deviation of the outdoor temperature from a reference temperature).

[0066] In some embodiments, before obtaining the fitness score corresponding to each initial control strategy, an initial abnormal control strategy that does not satisfy the preset expert rule is identified from the initial control strategies, the initial abnormal control strategy is removed, and the initial control strategy is regenerated until no initial control strategy is identified in the initial control strategy, and the first preset number is satisfied.

[0067] That is, the application discloses a screening method of a safe-oriented initial control strategy, that is, after the initial control strategy is acquired, the initial control strategies that do not satisfy preset expert rules are deleted by using the preset expert rules, and then the initial control strategies are scored in terms of fitness, so that the fitness of the initial control strategies is evaluated on the basis of effectively guaranteeing the safety of the genetic parent control strategies subjected to cross inheritance.

[0068] In step 302, a second preset number of initial control strategies are randomly extracted, and an initial control strategy with the highest fitness score is selected as a genetic parent control strategy, until a third preset number of genetic parent control strategies are acquired.

[0069] That is, in the embodiment of the application, in order to further improve the optimization efficiency of the control strategy, the application selects the initial control strategy with the highest fitness from the initial control strategies by using a random screening mechanism.

[0070] In some embodiments, the second preset number can be 3, that is, 3 initial control strategies can be randomly selected from a large number of initial control strategies, and then the fitness scores corresponding to the 3 initial control strategies are acquired, and the initial control strategy with the highest fitness score is selected as the genetic parent control strategy.

[0071] It should be understood that in the embodiment of the application, the third preset number of genetic parent control strategies can be repeated, that is, the genetic parent control strategies can be randomly screened with replacement, so as to effectively improve the fitness scores of the control strategies used for cross inheritance, and further improve the optimization efficiency of the control strategy optimization. In some embodiments, before the second preset number of initial control strategies are randomly extracted, the fitness scores of all the initial control strategies can be acquired, and then the first preset number of initial control strategies are selected from all the initial control strategies as the genetic parent control strategies, so as to improve the excellent proportion of the genetic parent control strategies and promote the evolution direction of the control strategy optimization.

[0072] In step 303, any two genetic parent control strategies are randomly extracted for cross inheritance, to obtain a genetic child control strategy corresponding to the genetic parent control strategies, until a fourth preset number of genetic child control strategies are acquired.

[0073] It should be noted that the cross inheritance is to exchange gene segments in the two genetic parent control strategies. The gene segment is part of the control scheme in the control strategy.

[0074] In some embodiments, in the process of randomly extracting any two genetic parent control strategies for cross inheritance to obtain a genetic child control strategy corresponding to the genetic parent control strategies, the two genetic parent control strategies are subjected to safe-oriented cross inheritance based on preset expert rules.

[0075] That is, in the embodiments of the present application, in addition to screening the control strategies for cross genetic in advance, the optimization process is further guided safely to improve the optimization efficiency of the control strategies.

[0076] In some embodiments, the safe guiding cross genetic of the two genetic parent control strategies includes: selecting any genetic parent control strategy from the two genetic parent control strategies, performing safety evaluation on each gene segment based on a preset expert rule to obtain a safety evaluation value corresponding to each gene segment of the genetic parent control strategy, selecting a cross point according to the safety evaluation value, and exchanging the gene segments of the two genetic parent control strategies according to the cross point to obtain two genetic child control strategies.

[0077] It should be noted that, in the embodiments of the present application, the safety evaluation on each gene segment based on the preset expert rule is specifically to judge the safety of the genetic child control model generated after the cross replacement of the two genetic parent control strategies from the first gene segment to the second gene segment.

[0078] It should be understood that the heating system is a complex system with multiple devices working together, for example, the valve needs to be closed when the water pump is closed, in other words, there is a dependent relationship between part of the control strategies. When the two genetic parent control strategies are cross replaced, a new unsafe factor may be generated in the genetic child control strategy after the cross replacement due to the control scheme in the other genetic parent control strategy. Therefore, in the embodiments of the present application, the safety of the cross replacement result is predicted before the cross replacement.

[0079] For example, the following formula can be used for safety evaluation:

[0080] wherein, is the safety evaluation value of the cross genetic of the two genetic parent control strategies from the first gene segment to the second gene segment, is a preset expert rule set , and is the two genetic parent control strategies from the first gene segment to the second gene segment.

[0081] Further, in some embodiments, selecting the cross point according to the safety evaluation value can include exchanging the gene segments by taking the highest α and β of the safety evaluation value as the cross point.

[0082] As described above, there is a dependency relationship between the partial control schemes, in order to further avoid ignoring the dependency relationship between the gene segments which slightly reduce the security evaluation value due to the replacement of other control schemes in the gene segment which increases the security evaluation value, the application further proposes: obtaining the dependency relationship between the gene segments according to the preset expert rules, determining the gene segment group with the dependency relationship and the gene position of each gene segment in the gene segment group according to the dependency relationship, and selecting the crossover point according to the gene position of each gene segment in the gene segment.

[0083] That is, the first gene segment and the second gene segment are compared with the gene position determined by the gene segment dependency relationship, if the replacement from the first gene segment to the second gene segment satisfies the dependency relationship between the gene segments, that is, at least two gene segments with the dependency relationship are not split, then the crossover genetic operation can be performed from the first gene segment to the second gene segment. If not, replace the crossover point and continue to evaluate.

[0084] Therefore, the application can always maintain the safety of the control strategy during the control strategy optimization process, effectively avoid generating a control strategy with unsafe factors, and perform safety guidance during the crossover genetic process. On the basis of ensuring the safety of the genetic offspring control strategy, the application effectively avoids the reduction of the control strategy in space and quantity caused by the control strategy elimination according to the preset expert rules after obtaining the genetic offspring control strategy, and ensures the reliability of the optimal control strategy search.

[0085] Step 304: obtaining the current offspring power consumption value corresponding to each genetic offspring control strategy of the current round, and performing the next round of crossover genetic operation until the difference between the offspring power consumption value corresponding to the current round and the offspring power consumption value corresponding to the previous round is less than a preset difference value.

[0086] It should be noted that the offspring power consumption value can be calculated according to the calculation method of the total power consumption described above, and the application will not be repeated here.

[0087] Specifically, after the third preset number of genetic offspring control strategies are obtained by cross breeding based on the third preset number of genetic parent control strategies, the current offspring power consumption values of the third preset number of genetic offspring control strategies are calculated, i.e., the total power consumptions corresponding to each genetic offspring control strategy, then the third preset number of genetic offspring control strategies of the current offspring are taken as the genetic parent control strategies of the next iteration round, the third preset number of genetic offspring control strategies of the next iteration round are obtained by cross breeding, and the total power consumptions corresponding to each genetic offspring control strategy of the third preset number of genetic offspring control strategies of the next iteration round are further calculated, then it is determined whether the difference between the total power consumptions corresponding to the genetic offspring control strategies of the two iteration rounds is less than a preset error, if not, the genetic control offspring control strategies of the next iteration round are continuously taken as the genetic parent control strategies of the next next iteration round for cross breeding until the difference between the total power consumptions of the genetic offspring control strategies of the two consecutive iteration rounds is less than the preset error.

[0088] It should be understood that the cross breeding process of the embodiment of the present application also includes preferentially selecting the genetic offspring control strategy with the lowest total power consumption as the genetic parent control strategy of the next iteration round in the iteration process to guarantee the demand for energy consumption optimization in the iteration process, which is not limited herein.

[0089] For example, the second preset number of genetic offspring control strategies can be randomly extracted in each iteration round, and the genetic offspring control strategy with the lowest total power consumption is selected as the genetic parent control strategy until the third preset number of genetic parent control strategies are obtained.

[0090] It should be noted that although the operations of the method of the present application are described in a specific order in the accompanying drawings, this does not require or imply that the operations must be performed in this specific order, or that all the shown operations must be performed to achieve the desired results.

[0091] Figure 4 A structural schematic diagram of a building energy efficiency optimization system provided by an embodiment of the present application is shown.

[0092] As shown in Figure 4 , the building energy efficiency optimization system 10 comprises: An acquisition module 11, configured to acquire historical load data, real-time running state data and meteorological data of a target time of a heating and ventilation system; The binuclear driving module 12 is configured to input the historical load data, the real-time operation state data and the meteorological data of the target time into a binuclear optimization model based on a Transformer-genetic algorithm and an expert rule to obtain an optimal control strategy; the optimal control strategy corresponds to the lowest power consumption; the binuclear optimization model comprises a Transformer-genetic algorithm sub-model and an expert rule sub-model; the Transformer-genetic algorithm sub-model is configured to predict a predicted load at the target time based on the historical load data, the real-time operation state data and the meteorological data of the target time, and optimize the control strategy based on the predicted load by using a genetic algorithm; the expert rule sub-model is configured to correct the control strategy in the optimization process based on a preset expert rule during the optimization of the control strategy by the Transformer-genetic algorithm sub-model based on the predicted load. The control module 13 is configured to control the heating and ventilation system to execute the optimal control strategy.

[0093] In some embodiments, the binuclear driving module 12 is specifically configured to: randomly generate a first preset number of initial control strategies, and obtain an adaptive score corresponding to each initial control strategy; randomly extract a second preset number of the initial control strategies, select the initial control strategy with the highest adaptive score as a genetic parent control strategy, and repeat the above process until a third preset number of genetic parent control strategies are obtained; randomly extract any two genetic parent control strategies to perform cross breeding to obtain a genetic child control strategy corresponding to the genetic parent control strategies, and repeat the above process until a fourth preset number of genetic child control strategies are obtained; obtain a child power consumption value corresponding to each genetic child control strategy in the current round, and perform cross breeding in the next round until the difference between the child power consumption value corresponding to the current round and the child power consumption value corresponding to the previous round is less than a preset error.

[0094] In some embodiments, the binuclear driving module 12 is specifically configured to: obtain the power consumption of the chiller, the chilled water pump, the cooling water pump and the cooling tower under the initial control strategy respectively, and calculate the total power consumption of the initial control strategy; calculate a first rule penalty factor and a second rule penalty factor corresponding to the initial control strategy according to the preset expert rule, and determine an expert rule penalty term according to the first rule penalty factor and the second rule penalty factor; determine the adaptive score according to the total power consumption and the expert rule penalty term.

[0095] In some embodiments, the dual-core driving module 12 is specifically configured to: In the process of performing cross genetic operations on the two genetic parent control strategies to obtain the genetic child control strategies corresponding to the genetic parent control strategies, the two genetic parent control strategies are subjected to safety-oriented cross genetic operations based on the preset expert rules.

[0096] In some embodiments, the dual-core driving module 12 is specifically configured to: Any one of the two genetic parent control strategies is selected, and safety evaluation is performed on each gene segment of the genetic parent control strategy based on the preset expert rules to obtain a safety evaluation value corresponding to each gene segment of the genetic parent control strategy; According to the safety evaluation value, a cross point is selected. According to the cross point, the two genetic parent control strategies are subjected to gene segment exchange to obtain the two genetic child control strategies.

[0097] In some embodiments, the dual-core driving module 12 is specifically configured to: According to the preset expert rules, a dependency relationship between the gene segments is obtained. According to the dependency relationship, a gene segment group having a dependency relationship and a gene position of each gene segment in the gene segment group are determined. According to the gene position of each gene segment in the gene segment group, the cross point is selected.

[0098] It should be understood that the modules or modules described in the building energy efficiency optimization system 10 correspond to the steps in the method described with reference to Figure 2 The operations and features described above with respect to the method are also applicable to the building energy efficiency optimization system 10 and the modules contained therein, and will not be described here. The building energy efficiency optimization system 10 can be pre- implemented in the browser or other security applications of the electronic device, or can be loaded into the browser or security applications thereof of the electronic device through downloading or the like. The corresponding modules in the building energy efficiency optimization system 10 can cooperate with the modules in the electronic device to realize the schemes of the embodiments of the present application.

[0099] In the foregoing detailed description, several modules or units mentioned are not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, a module or unit described above can be further divided into a plurality of modules or units to embody the features and functions thereof.

[0100] The following reference Figure 5 , Figure 5A structural diagram of a computer system of an electronic device or a server suitable for implementing the embodiments of the present application is shown, As Figure 5 shown, the computer system 500 includes a central processing unit (CPU) 501 which can perform various appropriate actions and processes in accordance with a program stored in a read only memory (ROM) 502 or a program loaded from a storage section 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data necessary for the operation instructions of the system are also stored. The CPU 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0101] The following components are connected to the I / O interface 505; an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as necessary. A removable media 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 510 as necessary, so that a computer program read out therefrom is installed into the storage section 508 as necessary.

[0102] In particular, the processes described above with reference to the flowcharts Figure 2 may be implemented as a computer software program. For example, the embodiments of the present application include a computer program product comprising a computer program carrying computer program code for executing the methods shown in the flowcharts. In such embodiments, the computer program comprises program code for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication section 509 and / or installed from the removable media 511. When the computer program is executed by the central processing unit (CPU) 501, the above-described functions defined in the system of the present application are performed.

[0103] It should be noted that the computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium or a combination thereof. The computer-readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a computer-readable storage medium can include, but are not limited to, the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer-readable storage medium can be any tangible medium that contains or stores a program used by an instruction execution system, apparatus, or device to function according to the program. In the present application, a computer-readable signal medium can include a computer-readable storage medium as described above, and a computer-readable signal medium can also be a computer-readable storage medium as described above in which case the computer-readable signal medium is a computer-readable storage medium that is also a computer-readable signal medium.

[0104] The flow diagrams and the block diagrams in the drawings are illustrations of possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flow diagrams or block diagrams can represent a module, a segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flow diagrams, and combinations of blocks in the block diagrams and / or flow diagrams, can be implemented by special purpose hardware-based systems that perform the specified functions or operations, or combinations of special purpose hardware and computer instructions.

[0105] As another aspect, the present application also provides a computer readable storage medium, which can be included in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable storage medium stores one or more programs, when the programs are used by one or more processors to execute a building energy efficiency optimization method described in the present application.

[0106] The above description is merely the preferred embodiments of the present application and the description of the technical principles used. It should be understood by those skilled in the art that the disclosed scope of the present application is not limited to the technical solutions formed by the specific combinations of the above technical features, and should also cover other technical solutions formed by the combinations of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the technical solutions formed by the mutual replacement of the above features and the technical features disclosed in the present application (but not limited to) having similar functions.

Claims

1. A building energy efficiency optimization method, characterized by, include: Acquire historical load data, real-time operating status data, and meteorological data for the target time of the HVAC system; The historical load data, the real-time operating status data, and the meteorological data at the target time are input into a dual-core optimization model constructed based on the Transformer-Genetic Algorithm and expert rules to obtain the optimal control strategy. The optimal control strategy corresponds to the lowest power consumption. The dual-core optimization model includes a Transformer-Genetic Algorithm sub-model and an expert rule sub-model. The Transformer-Genetic Algorithm sub-model is used to predict the predicted load at the target time based on the historical load data, the real-time operating status data, and the meteorological data at the target time, and uses a genetic algorithm to optimize the control strategy based on the predicted load. The expert rule sub-model is used to correct the control strategy during the optimization process using the genetic algorithm based on the predicted load, based on preset expert rules. Control the HVAC system to execute the optimal control strategy.

2. The building energy efficiency optimization method of claim 1, wherein, The method of optimizing the control strategy based on the predicted load using a genetic algorithm includes: A first preset number of initial control strategies are randomly generated, and the fitness score corresponding to each initial control strategy is obtained; Randomly extract a second preset number of the initial control strategies, and select the initial control strategy with the highest fitness score as the genetic parent control strategy, until a third preset number of the genetic parent control strategies are obtained. Randomly select any two of the genetic parent control strategies and perform cross-inheritance to obtain the genetic offspring control strategies corresponding to the genetic parent control strategies, until a fourth preset number of the genetic offspring control strategies are obtained. Obtain the power consumption value of the offspring corresponding to each of the genetic offspring control strategies in the current round, and perform the next round of cross-genesis until the difference between the power consumption value of the offspring in the current round and the power consumption value of the offspring in the previous round is less than a preset error.

3. The building energy efficiency optimization method according to claim 2, characterized in that, The step of obtaining the fitness score corresponding to each initial control strategy includes: The power consumption of the chiller, chilled water pump, cooling water pump and cooling tower under the initial control strategy is obtained respectively, and the total power consumption of the initial control strategy is calculated. Calculate the first rule penalty factor and the second rule penalty factor corresponding to the initial control strategy according to the preset expert rules, and determine the expert rule penalty item according to the first rule penalty factor and the second rule penalty factor; The fitness score is determined based on the total power consumption and the expert rule penalty.

4. The building energy efficiency optimization method according to claim 2, characterized in that, The process of modifying the control strategy during the optimization process based on preset expert rules includes: In the process of randomly extracting any two of the genetic parent control strategies and performing cross-genesis to obtain the genetic offspring control strategies corresponding to the genetic parent control strategies, the two genetic parent control strategies are subjected to safety-guided cross-genesis based on the preset expert rules.

5. The building energy efficiency optimization method according to claim 2 or 4, characterized in that, The safe-guided cross-inheritance of the two genetic parent control strategies includes: Select any one of the two genetic parent control strategies, and perform a safety assessment on a gene-segment-by-gene basis based on the preset expert rules to obtain the safety assessment value corresponding to each gene segment of the genetic parent control strategy; Based on the aforementioned safety assessment values, select the intersection point; Based on the crossover point, the two genetic parent control strategies are swapped to obtain two genetic offspring control strategies.

6. The building energy efficiency optimization method according to claim 5, characterized in that, The step of selecting the intersection point based on the security assessment value includes: According to the preset expert rules, the dependency relationships between the gene segments are obtained; Based on the dependency relationship, determine the gene segment group with the dependency relationship and the gene location of each gene segment in the gene segment group; The crossover point is selected based on the gene location of each gene segment in the gene segment group.

7. A building energy efficiency optimization system, characterized in that, include: The acquisition module is used to acquire historical load data, real-time operating status data, and meteorological data for the target time of the HVAC system. A dual-core drive module is used to input the historical load data, the real-time operating status data, and the meteorological data at the target time into a dual-core optimization model constructed based on Transformer-Genetic Algorithm and expert rules to obtain the optimal control strategy; the optimal control strategy corresponds to the lowest power consumption; wherein, the dual-core optimization model includes a Transformer-Genetic Algorithm sub-model and an expert rule sub-model; the Transformer-Genetic Algorithm sub-model is used to predict the predicted load at the target time based on the historical load data, the real-time operating status data, and the meteorological data at the target time, and uses a genetic algorithm to optimize the control strategy based on the predicted load; the expert rule sub-model is used to correct the control strategy during the optimization process by the Transformer-Genetic Algorithm sub-model using the genetic algorithm based on the predicted load based on preset expert rules; The control module is used to control the HVAC system to execute the optimal control strategy.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the building energy efficiency optimization method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the building energy efficiency optimization method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the building energy efficiency optimization method according to any one of claims 1-6.

Images