Method, apparatus, computer device, storage medium and program product of system operation optimization
Patent Information
- Application Number
- PCT/CN2025/085091
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-03-26
- Publication Date
- 2026-10-01
Smart Images

Figure CN2025085091_01102026_PF_FP_ABST
Abstract
Description
METHOD, APPARATUS, COMPUTER DEVICE, STORAGE MEDIUM AND PROGRAM PRODUCT OF SYSTEM OPERATION OPTIMIZATIONTECHNICAL FIELD
[0001] Embodiments of the application relate to the technical field of optimization, and particularly relates to a method, an apparatus, an electronic device, a storage medium and a program product of system operation optimization.BACKGROUND
[0002] Energy system is an essential part of industrial production process. Although energy systems are similar, the topology of the system and operation requirements vary a lot. Taking chilled water system as an example, the system usually consists of chillers, cooling towers and pumps. But the type of devices and how they connect to each other are decided by operation requirements on a case-by-case basis, e.g., number of chillers / pumps / cooling towers / water collectors, type of chillers (water-cooled, air-cooled) / pumps (water pump, heat pump) , several pumps are connected to one chiller group, or one pump is connected to one chiller, etc. Therefore, the solution for the energy system operation optimization needs to be customized from one case to another, which is time-consuming and not quite cost-effective.SUMMARY
[0003] The contents section of the present invention is provided to introduce in simplified form selected concepts which will be further described in the specific embodiments section below. The contents section of the invention is not intended to identify any key features or essential features of the claimed subject matter, nor is it intended to be used to assist in determining the scope of the claimed subject matter.
[0004] At least one of the embodiments of:
[0005] A method of system operation optimization, wherein, comprises: defining and generating a data class for said system operation based on topology data parsing; generating virtual environment according to configuration and said data class, and calling model with said virtual environment to produce first information; running said model to return multiple parameter sets of said system operation based on said first information.
[0006] By above means, it is enabled that easy to obtain optimized parameter sets of system operation for optimized controlling.
[0007] The application also discloses an apparatus of system operation optimization, wherein, comprising:
[0008] data class module, for defining and generating a data class for said system operation based on topology data parsing;
[0009] virtual environment module, for generating virtual environment according to configuration and said data class and calling model with said virtual environment to produce first information;
[0010] model running module, for running said model to return multiple parameter sets of said system operation based on said first information.
[0011] The application also provides a computer device comprising a memory and a processor, said memory storing a computer program, wherein said processor implements above said method when said computer program is executed by said processor.
[0012] The application also provides a computer readable storage medium having a computer program stored thereon, wherein said computer program implements above said method when executed by the processor.
[0013] The application also provides a computer program product, said computer program product being tangibly stored on a computer-readable medium and comprising computer-executable instructions, said computer-executable instructions when executed causing at least one processor to perform above said method as described above.DESCRIPTION OF DRAWINGS
[0014] To more clearly describe technical solutions in embodiments of the present disclosure or the prior art, drawings to be used in the description of the embodiments or the prior art will be briefly introduced below. Apparently, the drawings in the description below are merely some embodiments disclosed in the embodiments of the present disclosure. For those of ordinary skills in the art, other drawings may also be obtained based on these drawings.
[0015] FIG. 1 is schematic flowchart of a method of system operation optimization provided in an embodiment of the present disclosure.
[0016] FIG. 2 is a schematic diagram of an apparatus of system operation optimization provided in an embodiment of the present disclosure.
[0017] FIG. 3 is a schematic diagram of a computer device of system operation optimization provided in an embodiment of the present disclosure.
[0018] FIG. 4 is a schematic diagram of a chilled water system provided in an embodiment of the present disclosure.
[0019] FIG. 5 is a schematic diagram of current optimization workflow provided in an embodiment of the present disclosure.
[0020] FIG. 6 is a schematic diagram of proposed optimization workflow provided in an embodiment of the present disclosure.
[0021] List of reference numerals:
[0022] S101-S103: method steps
[0023] 200: apparatuses
[0024] 201: a waiter service module
[0025] 202: a count service module
[0026] 203: an artificial intelligence farm module
[0027] 300: computer device
[0028] 302: processor
[0029] 304: memoryDETAILED DESCRIPTION
[0030] In the following specification, many specific details are set forth for explanatory purposes. However, it will be appreciated that the realization of the present invention can be carried out without these specific details. In other examples, well-known circuits, structures, and techniques are not shown in detail so as not to affect the understanding of the specification.
[0031] In previous, developing a solution for energy system operation optimization mainly consists of following steps, as show in Fig. 5:
[0032] Data scientists talk to the customers and the digitalization engineers, define the optimization problem incl. observation variables, set points, target, constraints, etc.
[0033] Data scientists or domain experts build models for system dynamic representation. The models are used for operation strategy / action exploration. For example, given a state and an action, the models return next state, reward, cost, whether the constraints are violated, etc.
[0034] Data scientists use optimization algorithms to train a strategy (e.g., given a state, the strategy tells the action) or explore the optimal action (e.g., given a state, find an action which satisfies the constraints, maximize the reward, minimize the cost, etc. ) .
[0035] This application proposes a framework to solve the problem through generic and modularized optimization workflow. Furthermore, the proposed framework breaks current optimization workflow into several configurable modules. The customized solution can be developed by putting these modules together per requirement.
[0036] In details, this application provides a method of system operation optimization, wherein, comprises:
[0037] S101, defining and generating a data class for said system operation based on topology data parsing.
[0038] In some embodiment, defining and generating a data class for said system operation based on topology data parsing, comprising:
[0039] defining and generating a data class including observation space, action space, objectives and constraints for said system operation based on topology data parsing.
[0040] In some embodiment, define a generic data class as below shows that contains the necessary information for system operations.
[0041] class DataClass:
[0042] Then, parse topology data and automatically generate a set of data class which defines observation space, action space, objectives, constraints, etc as below shows:
[0043] <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CA0F10>, 'Pump_1' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CA0F40>, 'Pump_2' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CEEDC0>, 'Pump_3' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CEEE50>, 'Pump_4' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC62E0>, 'Pump_5' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6430>, 'Pump_6' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6400>, 'Pump_8' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6490>, 'Pump_7' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6460>, 'Chiller_1' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC64F0>, 'CoolingTower_1' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC64C0>, 'CoolingTower_2' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6550>, 'CoolingTower_3' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6520>, 'Chiller_2' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC65B0>, 'Chiller_3' : <WFC_extension_on_AUS. WFC_optimization_modules. dataClasses. DataClass object at 0x0000025148CC6580>} .
[0044] S102, generating virtual environment according to configuration and said data class, and calling model with said virtual environment to produce first information.
[0045] In some embodiment, generating virtual environment according to configuration and said data class, comprising:
[0046] generating virtual environment according to configuration, observation space, action space and constraint.
[0047] In some embodiment, automatically generate a virtual environment for optimal operation strategy learning and action exploration according to the configuration and the data class defined in above including. observation space, action space, constraint, etc as below shows:
[0048] <dcEnv instance>
[0049] observation space: Dict ( 'Chiller_1. COP' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Chilled_Water_Inlet_Mass_Flow_Rate' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Chilled_Water_Inlet_Pressure' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Chilled_Water_Inlet_Temperature' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Chilled_Water_Outlet_Mass_Flow_Rate' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Chilled_Water_Outlet_Pressure' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Chilled_Water_Outlet_Temperature' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Condenser_Pressure_Ratio1' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Capacity' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Water_Inlet_Mass_Flow_Rate' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Water_Inlet_Pressure' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Water_Inlet_Temperature' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Water_Outlet_Mass_Flow_Rate' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Water_Outlet_Pressure' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Cooling_Water_Outlet_Temperature' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Evaporator_Pressure_Ratio1' : Box (0.0, 1.0, (1, 1) , float32) , 'Chiller_1. Evaporator_Pressure_Ratio2' : Box (0.0, 1.0, (1, 1) , float32) ……
[0050] …action space: Dict ( 'Chiller_1. On_Off_Status' : MultiBinary (4) , 'Chiller_1. Realtime_Power' : Box (0.0, 1.0, (1, 4) , float32) , 'Chiller_2. On_Off_Status' : MultiBinary (4) , 'Chiller_2. Realtime_Power' : Box (0.0, 1.0, (1, 4) , float32) , 'Chiller_3. On_Off_Status' : MultiBinary (4) , 'Chiller_3. Realtime_Power' : Box (0.0, 1.0, (1, 4) , float32) , 'CoolingTower_1. Fan_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'CoolingTower_1. Spray_Water_Motor_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'CoolingTower_2. Fan_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'CoolingTower_2. Spray_Water_Motor_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'CoolingTower_3. Fan_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'CoolingTower_3. Spray_Water_Motor_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'Pump_1. Motor_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'Pump_2. Motor_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'Pump_3. Motor_Speed' : Box (0.0, 1.0, (1, 4) , float32) , 'Pump_3. On_Off_Status' : MultiBinary (4) , 'Pump_6. On_Off_Status' : MultiBinary (4) )
[0051] {'CoolingConsumptionEquipment' : {'Expected_Cooling_Capacity' : array ( [ [1000., inf] ,
[0052] [1000., inf] ,
[0053] [1000., inf] ,
[0054] [1000., inf] ] ) } }
[0055] In some embodiment, the virtual environment is a class. The environment can be reset with current system state. Observations and actions can be randomly sampled from the observation space and action space of the environment. Given current state and one action, the environment calls the system dynamic model (domain model, or simulation model, or data-driven model) to get the information incl. next state, KPI, constraint violations, etc.
[0056] In more detail, in this application, the code is just an example for implementation design by python. In the code, this application just use python function or python function set for implementation and realization. The similar functionality or method could be implemented by any code type such as Json, Jave or others. This application does not have restrictions on this.
[0057] S103, running said model to return multiple parameter sets of said system operation based on said first information.
[0058] In some embodiment, automatically tune hyper-parameters of the selected optimization algorithms with configurations as below shows:
[0059] recommendation, recommendation_index, stats, actions, logs, violation_details =callCmaesAgent (dataclass=dc, datainstance=datainstance, env=env, timelimit=ConfigOptimization [header. TIMEOUT] , config=ConfigCMAES, logger=logger)
[0060] And in some embodiment, running through the defined search space (e.g., lower bound, upper bound, step size, etc. ) with the training and testing data; and further return the top N hyper-parameter sets regarding KPIs such as cost, reward, computation time, etc. In some embodiment, the first information could include KPIs and constraint violation such as cost, reward, computation time.
[0061] In some embodiment, this application could invoke or call different type of model for running and implementing to realize the method or functionality of this solution. Not have restrictions on model calling or model realization.
[0062] The technical effect of this application compared with the current solution has the following:
[0063] Currently a tailored solution usually relies on at least three groups of people, i.e., experts from customers who know system operation requirements, digitalization engineers who are familiar with IT to OT implementation, and data scientists who perform data analysis and algorithm development. Due to knowledge gap between different groups of people, the solution development takes lots of communications and manual work. With the proposed framework, digitalization engineers could develop a customized solution alone without involving data scientist as Fig. 6 shows.
[0064] The way to detect whether this application being used could follow next steps:
[0065] 1. are the optimization workflow and the modules automatically adaptable to various energy systems and devices; 2. can the optimization workflow be built with modules; 3. Do data scientist need to be involved.
[0066] Although the individual steps in the flowchart of FIG. 1 are shown sequentially as indicated by the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless expressly stated herein, there is no strict order limitation on the execution of these steps, and the steps may be executed in other orders. Moreover, at least a portion of the steps of FIG. 1 may include multiple steps or multiple stages, which are not necessarily executed to completion at the same moment, but may be executed at different moments, and the order in which these steps or stages are executed is not necessarily sequential, but may be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.
[0067] As shown in FIG. 2, the application provides an apparatus of system operation optimization, wherein, comprising:
[0068] data class module 201, for defining and generating a data class for said system operation based on topology data parsing;
[0069] virtual environment module 202, for generating virtual environment according to configuration and said data class and calling model with said virtual environment to produce first information;
[0070] model running module 203, for running said model to return multiple parameter sets of said system operation based on said first information.
[0071] It is noted that the device may comprise more or fewer modules to perform the described functions. For example, at least one of the modules of FIG. 2 may be further divided into plural distinct sub-modules, each of which is used to perform at least a portion of the operations described herein in conjunction with the corresponding module. In addition, in some examples, the device 200 may include additional modules for performing other operations already described in the specification. In addition, it will be understood by those skilled in the art that the exemplary device 200 may be implemented with software, hardware, firmware, or any combination thereof.
[0072] FIG. 3 provides a computer device. According to one embodiment, the computer device 300 may include a processor 302, the processor 302 executing a computer program stored in a memory 304. The computer program is executed by the processor to implement the method described above.
[0073] It will be understood by one of ordinary skill in the art that the structure illustrated in FIG. 3, which is only a block diagram of a portion of the structure related to the embodiments of the present application, does not constitute a limitation on the computer device to which the present application is applied, and that a specific computer device may include more or fewer components than those shown in the drawings, or a combination of some of the components, or have a different arrangement of components.
[0074] A person of ordinary skill in the art may understand that all or part of the processes in the methods for realizing the above embodiments are possible to be accomplished by a computer program for instructing the relevant hardware, and that said computer program may be stored in a non-volatile computer-readable storage medium, which computer program, when executed, may comprise processes such as the processes of the embodiments of each of the above-described methods. Among other things, any reference to a memory, storage, database, or other medium used in the various embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include Read-Only Memory (ROM) , magnetic tape, floppy disk, flash memory, or optical memory. Volatile memory may include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, the RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM) , and the like.
[0075] The present application also provides a computer readable storage medium having a computer program stored thereon, said computer program realizing the above steps when executed by a processor.
[0076] The present application also provides a computer program product, said computer program product being tangibly stored on a computer-readable medium and comprising computer-executable instructions, said computer-executable instructions when executed causing at least one processor to perform said method.
[0077] Further, said computer program may be stored, run in the cloud for execution of said method. Further, components of said program may be laid out on multiple devices, on the cloud, e.g. the corresponding steps may be laid out, run on a local or local computer, or run on different cloud devices, transmitting signals via a communication connection, or may also be laid out, run on a local or local computer. The present application does not limit the described ways or methods, and the corresponding techniques can be flexibly laid out and deployed to fully utilize the cloud, big data, supercomputing power, and other devices and techniques for the execution and completion of the methods.
[0078] Some implementations of the present disclosure may include artifacts. The artifacts may include a storage medium, which is used to store logic. Examples of storage media may include one or more types of computer-readable storage media capable of storing electronic data, including volatile memory or non-volatile memory, removable or non-removable memory, erasable or non-erasable memory, writable or rewritable memory, and the like. Examples of logic may include various software units, such as software components, programs, applications, computer programs, application programs, system programs, machine programs, operating system software, middleware, firmware, software modules, routines, subroutines, functions, methods, procedures, software interfaces, application program interfaces (APIs) , instruction sets, computational code, computer code, code segments, computer code segments, words, value, symbol, or any combination thereof. In some implementations, for example, the article may store executable computer program instructions that, when executed by the processor, cause the processor to perform the methods and / or operations described herein. The executable computer program instructions may include any suitable type of code, e.g., source code, compiled code, interpreted code, executable code, static code, dynamic code, and the like. The executable computer program instructions may be implemented according to a predefined computer language, manner, or syntax for commanding a computer to perform a particular function. Said instructions may be implemented using any suitable high-level, low-level, object-oriented, visual, compiled, and / or interpreted programming language.
[0079] Throughout the specification, there are references to “an embodiment” , “realization” , “exemplary embodiment” , “some embodiments” , “various embodiments” , “various embodiments” , “various embodiments” , “various embodiments” , “various embodiments” , and “various embodiments” , References to “an implementation” , “implementation” , “exemplary implementation” , “some implementations” , “various implementations” , etc., throughout the specification indicate that the described implementations of the present invention may include particular features, structures, or characteristics, however, it is not necessary for each implementation to include these particular features, structures, or characteristics. In addition, some implementations may have some, all, or none of the features described with respect to other implementations.
[0080] What has been described above includes examples of the disclosed architecture. It is certainly not possible to describe every conceivable combination of components and / or methods, but those skilled in the art can appreciate that many other combinations and arrangements are possible. Accordingly, the novel architecture is intended to cover all such substitutions, modifications, and variations that fall within the spirit and scope of the appended claims.
[0081] Independent of the grammatical term usage, individuals with male, female or other gender identities are included within the term.
Claims
1.A method of system operation optimization, wherein, comprises:defining and generating a data class for said system operation based on topology data parsing;generating virtual environment according to configuration and said data class, and calling model with said virtual environment to produce first information;running said model to return multiple parameter sets of said system operation based on said first information.2.The method according to claim 1, wherein defining and generating a data class for said system operation based on topology data parsing, comprising:defining and generating a data class including observation space, action space, objectives and constraints for said system operation based on topology data parsing.3.The method according to any one claims 1 to 2, wherein generating virtual environment according to configuration and said data class, comprising:generating virtual environment according to configuration, observation space, action space and constraint.4.The method according to any one claims 1 to 3, wherein generating virtual environment according to configuration and said data class, comprising:resetting said virtual environment with current system state.5.The method according to any one claims 1 to 4, wherein calling model with said virtual environment to produce first information, comprising:sampling observation and action from said observation space and said action space.6.The method according to any one claim 1 to 5, wherein calling model with said virtual environment to produce first information, comprising:calling model with said virtual environment to produce first information including KPI and constraint violation.7.The method according to any one claim 1 to 6, wherein running said model to return multiple parameter sets of said system operation based on said first information, comprising:running through search space with training and testing data.8.An apparatus (200) of system operation optimization, wherein, comprises:data class module, for defining and generating a data class for said system operation based on topology data parsing;virtual environment module, for generating virtual environment according to configuration and said data class and calling model with said virtual environment to produce first information;model running module, for running said model to return multiple parameter sets of said system operation based on said first information.9.A computer device comprising a memory and a processor, said memory storing a computer program, wherein said processor realizes the steps of the method described in any one of claims 1 to 7 when said computer program is executed by said processor.10.A computer readable storage medium, wherein having stored a computer program, said computer program executed by the processor implements the steps of any one of claims 1 to 7.11.A computer program product, said computer program product being tangibly stored on a computer-readable medium and comprising computer-executable instructions, said computer-executable instructions executed causing at least one processor to perform the method according to any one of claims 1 to 7.