WO2023179076A1 - 基于混合整数规划的针对工业设施的负荷分解方法和装置 - Google Patents

基于混合整数规划的针对工业设施的负荷分解方法和装置 Download PDF

Info

Publication number
WO2023179076A1
WO2023179076A1 PCT/CN2022/135019 CN2022135019W WO2023179076A1 WO 2023179076 A1 WO2023179076 A1 WO 2023179076A1 CN 2022135019 W CN2022135019 W CN 2022135019W WO 2023179076 A1 WO2023179076 A1 WO 2023179076A1
Authority
WO
WIPO (PCT)
Prior art keywords
equipment
load
industrial
dynamic
dynamic load
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2022/135019
Other languages
English (en)
French (fr)
Inventor
李楚一
郑可迪
郭鸿业
陈启鑫
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
Original Assignee
Tsinghua University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tsinghua University filed Critical Tsinghua University
Publication of WO2023179076A1 publication Critical patent/WO2023179076A1/zh
Priority to US18/759,982 priority Critical patent/US20240364107A1/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for AC mains or AC distribution networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for AC mains or AC distribution networks
    • H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2103/00—Details of circuit arrangements for mains or AC distribution networks
    • H02J2103/30—Simulating, planning, modelling, reliability check or computer assisted design [CAD] of electric power networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Definitions

  • the present application relates to the technical field of non-intrusive load decomposition, and in particular to a load decomposition method and device for industrial facilities based on mixed integer programming.
  • load decomposition is usually divided into three types: load decomposition in civil, commercial and industrial settings.
  • load decomposition in civil, commercial and industrial scenarios.
  • load decomposition has more significant profit potential (the electricity consumption of one factory may be equivalent to that of hundreds of residences)
  • current research on load decomposition has mainly focused on residential facilities.
  • civil scenarios industrial scenarios have a large number of dynamic load devices and the power consumption characteristics are more complex, which brings great difficulties to load decomposition.
  • NILM non-intrusive load monitoring, non-intrusive power load monitoring
  • related technologies first studied load decomposition in large industrial buildings by collecting power data from a large industrial cold storage and combining the collected data with Typical residential data were compared and analyzed, and two benchmark models, CO (Combinationrial Optimization, combination optimization algorithm) and FHMM (Factory Hidden Markov Models, multiple hidden Markov models) were used to decompose the industrial load, and proposed solutions for Improvement measures for sexual training and sub-measurement.
  • Industrial loads have characteristics such as a large proportion of dynamic loads and assembly line dependence that are rare in civilian loads. These characteristics will greatly affect the effects of traditional optimization models and FHMM-based models. Relevant research using deep learning methods shows that deep learning methods can still achieve good results in industrial scenarios. At the same time, these models are highly dependent on a large amount of training data, but industrial data are often very sensitive, which has become a major problem in model application. obstacle.
  • k-means clustering technology This technology can iteratively divide the data set into k categories according to distance given the number of cluster centers k.
  • the solution to the above relaxation problem is the optimal solution in a larger feasible domain.
  • the solution to the original problem must not be better than this solution, so the solution to the relaxation problem is defined as a "lower bound" of the original problem.
  • the solution to the relaxation problem is defined as a "lower bound" of the original problem.
  • branching sub-problems will continue to be divided. It is possible to obtain the optimal solution that satisfies all constraints of the sub-problem. This solution is only the optimal within the local feasible domain of the original problem and is not necessarily the global optimal. Therefore, the solution The solution is specified as an "upper bound" on the original problem.
  • the branch and bound method is an iterative algorithm. As the sub-problems are solved, the upper and lower bounds are continuously updated. When the algorithm meets the convergence conditions, the numerical solution to the original problem can be obtained.
  • solvers are generally used to solve such problems. These solvers generally use many algorithms to optimize the solving speed, such as the cutting plane method, etc.
  • This application provides a load decomposition method, device, electronic equipment and storage medium for industrial facilities based on mixed integer programming to solve the problem that existing load decomposition technology is difficult to effectively process industrial load data under the constraints of insufficient industrial load data and difficulty in data acquisition.
  • load characteristics of dynamic load equipment and stable load equipment can be considered at the same time, and non-intrusive load decomposition with high accuracy can be carried out in industrial scenarios.
  • the first embodiment of the present application provides a load decomposition for industrial facilities based on mixed integer programming, which includes the following steps:
  • the industrial equipment to be decomposed is divided into dynamic load equipment and stable load equipment according to power consumption characteristics, and the equipment-level load curve of the dynamic load equipment is digitally filtered, and time invariance is used to expand the pulse part of the load that fluctuates frequently. , obtain the load curve of the processed dynamic load equipment;
  • a mixed integer programming model is constructed based on the processed dynamic load equipment load curve and the stable load equipment load curve. After introducing the correction of industrial process constraints and equipment operation constraints into the mixed integer programming model, the optimal variable optimization variable is obtained by solving the problem. the best value;
  • a decomposition result is reconstructed for the processed dynamic load equipment combination signal matrix, and a decomposition result is reconstructed for converting the stable load equipment into a power state sequence based on the optimization problem model.
  • a processed dynamic load device is obtained based on the new basis vector and the violently fluctuating pulse part.
  • the industrial equipment to be decomposed is divided into dynamic load equipment and stable load equipment according to power consumption characteristics, including:
  • the stable load equipment is modeled using a combination optimization method:
  • n is the device number
  • b n,k (t) is the variable that determines the power state of device n
  • k is [1, K The integer in n ]
  • K n is the number of states of equipment n; the dynamic load equipment is modeled by a linear combination of the known equipment-level load curves of each equipment:
  • D n is the signal matrix of device n
  • a n is the activation coefficient matrix
  • constructing a mixed integer programming model based on the processed dynamic load equipment load curve and the stable load equipment load curve also includes:
  • Power state matrix B n is the state selection matrix
  • m is the time series length of an optimized interception.
  • the shortest duration w n,k is constrained by a set of linear inequalities:
  • ⁇ is the regularization coefficient
  • R is the process restriction matrix
  • the second aspect embodiment of the present application provides a load decomposition device for industrial facilities based on mixed integer programming, including:
  • the acquisition module is used to obtain the equipment list and the load curve to be decomposed of the industrial equipment to be decomposed, and obtain the equipment-level load curve of each equipment based on the equipment list and the load curve to be decomposed;
  • a processing module configured to divide the industrial equipment to be decomposed into dynamic load equipment and stable load equipment according to power consumption characteristics, digitally filter the equipment-level load curve of the dynamic load equipment, and utilize time invariance to expand the load center For the pulse part that fluctuates frequently, the processed load curve of the dynamic load equipment is obtained;
  • a first calculation module configured to construct a mixed integer programming model based on the processed dynamic load equipment load curve and the stable load equipment load curve, and introduce corrections to industrial process constraints and equipment operation limitations in the mixed integer programming model. Finally, solve to obtain the optimal value of the optimization variable;
  • the second calculation module is used to reconstruct the decomposition result of the processed dynamic load equipment combination signal matrix based on the optimization problem model, and reconstruct the stable load equipment into a power state sequence based on the optimization problem model. Break down the results.
  • processing module is specifically used for:
  • a processed dynamic load device is obtained based on the new basis vector and the violently fluctuating pulse part.
  • processing module is used for:
  • the stable load equipment is modeled using a combination optimization method:
  • n is the power of device n at time t
  • t is the time
  • T is the time series length of an optimization
  • n is the device number
  • b n,k (t) is the variable that determines the power state of device n
  • k is [1, K n ] is an integer
  • K n is the number of states of device n;
  • the dynamic load equipment is modeled by a linear combination of the known equipment-level load curves of each equipment:
  • D n is the signal matrix of device n
  • a n is the activation coefficient matrix
  • the first computing module is also used to:
  • Power state matrix B n is the state selection matrix
  • m is the time series length of an optimized interception.
  • the first computing module is also used to:
  • the shortest duration w n,k is constrained by a set of linear inequalities:
  • ⁇ is the regularization coefficient
  • R is the process restriction matrix
  • a third embodiment of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor.
  • the processor executes the program to implement The load decomposition method for industrial facilities based on mixed integer programming as described in the above embodiments.
  • the fourth embodiment of the present application provides a computer-readable storage medium on which a computer program is stored, and the program is executed by a processor to implement the above-mentioned load decomposition method for industrial facilities based on mixed integer programming.
  • the industrial equipment is classified into one type: dynamic load equipment with continuously adjustable power, and the other type: stable load equipment with power switching between fixed states; equipment levels of a certain scale
  • the electricity consumption data is used as a training set, and the digital filtering method is used to process the appropriate dynamic load equipment load curve, and the time invariance of the pulse part of the load that fluctuates frequently is used to expand its training data set; the equipment information and training data are used to construct a decomposition Mixed integer programming model; use physical constraints such as industrial process dependencies of some equipment to introduce model correction terms or constraints; solve planning problems to obtain decomposition results.
  • Figure 1 is a flow chart of a load decomposition method for industrial facilities based on mixed integer programming provided according to an embodiment of the present application
  • Figure 2 is a flow chart of a load decomposition method for industrial facilities based on mixed integer programming provided according to an embodiment of the present application
  • Figure 3 is a block diagram of a load decomposition device for industrial facilities based on mixed integer programming provided according to an embodiment of the present application
  • Figure 4 is a schematic diagram of an electronic device according to an embodiment of the present application.
  • this application provides a method based on mixed integer programming A load decomposition method for industrial facilities.
  • the industrial equipment is classified into one category for dynamic load equipment with continuously adjustable power and one category for power between fixed states. Switching stable load equipment; and equipment-level power consumption data of a certain scale are used as training sets.
  • Digital filtering methods are used to process appropriate dynamic load equipment load curves, and the time invariance of the frequently fluctuating pulse part of the load is used to expand the training data.
  • Set use equipment information and training data to build a mixed integer programming model for decomposition; use physical limitations such as industrial process dependencies of some equipment to introduce model correction terms or constraints; solve the planning problem to obtain the decomposition results.
  • FIG. 1 is a schematic flow chart of a load decomposition method for industrial facilities based on mixed integer programming provided by an embodiment of the present application.
  • the load decomposition method for industrial facilities based on mixed integer programming includes the following steps:
  • step S101 an equipment list of the industrial equipment to be decomposed and a load curve to be decomposed are obtained, and an equipment-level load curve of each equipment is obtained based on the equipment list and the load curve to be decomposed.
  • the power reading sequence recorded by the user's low-voltage circuit smart meter in a day is defined as the "total load curve” and recorded as a column vector; it is defined that separate sensors are installed on the electrical equipment, and the power reading sequence collected for the day is "equipment level”.
  • Load curve recorded as a column vector of the same dimension.
  • the equipment asset list of the industrial facilities to be decomposed is obtained; the load curve to be decomposed is obtained from the smart meter of the industrial facility to be decomposed, and the equipment-level load curve of each equipment is obtained from the equipment manufacturer or the equipment load database.
  • step S102 the industrial equipment to be decomposed is divided into dynamic load equipment and stable load equipment according to the power consumption characteristics, and the equipment-level load curve of the dynamic load equipment is digitally filtered, and time invariance is used to expand frequently fluctuating pulses in the load. Part, the processed load curve of the dynamic load equipment is obtained.
  • dividing the industrial equipment to be decomposed into dynamic load equipment and stable load equipment according to power consumption characteristics includes: modeling the stable load equipment using a combined optimization method:
  • n is the device number
  • n is the device number
  • the power value of each state is determined using k-means clustering technology.
  • b n, k (t) is the variable that determines the power state of device n.
  • k is an integer in [1, K n ].
  • K n is the value of device n. Number of states;
  • Dynamic load equipment is modeled using a linear combination of the known equipment-level load curves of each equipment:
  • D n is the signal matrix of device n
  • each of its column vectors is the known device-level load curve of the device, which is called a basis vector
  • a n is the activation coefficient matrix, that is, the basis vector
  • the equipment-level load curve of the dynamic load equipment is digitally filtered, and time invariance is used to expand the frequently fluctuating pulse part of the load to obtain the processed dynamic load equipment, including:
  • the device-level load curve of the device first uses median filtering to remove noise spikes, then detects and separates large slope parts in the device-level load curve of dynamic load devices, and finally decomposes the load of each dynamic device into smooth basic parts and fluctuations Violent pulse part; generate a column vector whose elements in the preset time period are all 0, connect the column vector to the beginning of each base vector to be expanded, and discard the sequence of the same length at the end of the base vector to be expanded, resulting in a lag of one unit
  • the expanded pulse is superimposed on the smooth basic part to obtain a new basis vector; the processed dynamic load equipment is obtained based on the new basis vector and the violently fluctuating pulse part.
  • industrial equipment is classified according to the power consumption characteristics of the equipment when it is working, and the equipment is divided into dynamic load equipment and stable load equipment; dynamic load equipment generally includes adjustable speed motors, and its load curve changes in a continuous range; stable load equipment The load curve switches between a fixed series of power states; the two types of equipment are modeled in different ways.
  • the equipment-level load curve of dynamic load equipment use digital filtering method to process the curve vector corresponding to the appropriate equipment, and use the time invariance of the frequently fluctuating pulse part in the load to construct a new realistic vector, thereby expanding the training set; for dynamic
  • the equipment-level load curve of the load equipment is first digitally filtered; during filtering, the median filter is first used to remove the noise peak, and then the large slope part in the curve is detected to separate them.
  • the limiter is added to assist the identification of pulses, and finally each The load of the equipment is decomposed into a smooth basic part and a violently fluctuating pulse part.
  • step S103 a mixed integer programming model is constructed based on the processed dynamic load equipment load curve and stable load equipment load curve. After introducing the correction of industrial process constraints and equipment operation constraints into the mixed integer programming model, the optimal optimization variable is obtained by solving Take value.
  • the diagonal block of Q is 1, the dimension corresponds to the number of grouped basis vectors, other area elements are 0, A is the activation coefficient matrix, N is the number of stable consistent devices, X n is the power state matrix, B n is the state selection matrix, and m is the time series length of an optimized interception.
  • modifications to the industrial process constraints and equipment operating limitations are introduced in the mixed integer programming model, including: the process constraints are adding additional penalty terms to the optimization objective function (mixed integer programming model):
  • the shortest duration w n,k is constrained by a set of linear inequalities:
  • ⁇ is the regularization coefficient
  • R is the process restriction matrix.
  • R is connected by a series of unit matrices.
  • the corresponding blocks are positive unit matrices and negative unit matrices.
  • step S104 a decomposition result is reconstructed based on the optimization problem model for the processed dynamic load equipment combination signal matrix, and a decomposition result is reconstructed based on the optimization problem model for converting the stable load equipment into a power state sequence.
  • mixed integer programming solving technology is used to solve the above optimization problem to obtain A, B n , the dynamic load equipment load is reconstructed through the load model DA, and the wind load equipment load is reconstructed through B n X n , thereby obtaining the decomposition result.
  • this application combines the advantages of the classic combination optimization model when dealing with stable load equipment and the advantages of the matrix decomposition model when dealing with dynamic load equipment, and solves it uniformly in an optimization planning problem. Compared with other methods based on probability models and The algorithm based on the optimization model is more suitable for complex industrial scenarios where different load types are mixed. On the other hand, this application also considers the problem of data dependence. Compared with widely used deep learning models, this application can achieve the same level of decomposition effect as general deep learning models while using less training data. This application can also make full use of existing data in practical industrial application scenarios to obtain more accurate load decomposition results, which has important significance and good application prospects.
  • Figure 2 is a flow chart of this load decomposition method for industrial facilities based on mixed integer programming.
  • the industrial equipment is classified into one type of dynamic load equipment with continuously adjustable power and one type of equipment. It is a stable load equipment with power switching between fixed states; a certain scale of equipment-level power consumption data is used as a training set, and the digital filtering method is used to process the appropriate dynamic load equipment load curve, and the time-varying pulse part of the load fluctuates frequently.
  • Figure 3 is a block diagram of a load decomposition device for industrial facilities based on mixed integer programming according to an embodiment of the present application.
  • the load decomposition device 10 for industrial facilities based on mixed integer programming includes: an acquisition module 100 , a processing module 200 , a first calculation module 300 and a second calculation module 400 .
  • the acquisition module 100 is used to obtain the equipment list and the load curve to be decomposed of the industrial equipment to be decomposed, and obtain the equipment-level load curve of each equipment based on the equipment list and the load curve to be decomposed;
  • the processing module 200 is used to divide the industrial equipment to be decomposed into dynamic load equipment and stable load equipment according to power consumption characteristics, digitally filter the equipment-level load curve of the dynamic load equipment, and use time invariance to expand frequently fluctuating parts of the load. In the pulse part, the processed load curve of the dynamic load equipment is obtained;
  • the first calculation module 300 is used to construct a mixed integer programming model based on the processed dynamic load equipment load curve and stable load equipment load curve. After introducing the correction of industrial process constraints and equipment operation constraints into the mixed integer programming model, the solution is optimized. The optimal value of the variable; and
  • the second calculation module 400 is used to reconstruct the decomposition result of the processed dynamic load equipment combination signal matrix based on the optimization problem model, and reconstruct the decomposition result of converting the stable load equipment into a power state sequence based on the optimization problem model.
  • the processing module 200 is specifically used to:
  • the processed dynamic load equipment is obtained based on the new basis vector and the violently fluctuating pulse part.
  • processing module 200 is also used to:
  • the stable load equipment is modeled using a combination optimization method:
  • n is the power of device n at time t
  • t is the time
  • T is the time series length of an optimization
  • n is the device number
  • b n,k (t) is the variable that determines the power state of device n
  • k is [1, K n ] is an integer
  • K n is the number of states of device n;
  • Dynamic load equipment is modeled using a linear combination of the known equipment-level load curves of each equipment:
  • D n is the signal matrix of device n
  • a n is the activation coefficient matrix
  • the first calculation module 300 is also used to:
  • Power state matrix B n is the state selection matrix
  • m is the time series length of an optimized interception.
  • the first calculation module 300 is also used to:
  • the shortest duration w n,k is constrained by a set of linear inequalities:
  • ⁇ is the regularization coefficient
  • R is the process restriction matrix
  • the industrial equipment is classified into one type of dynamic load equipment with continuously adjustable power and one type of equipment. It is a stable load equipment with power switching between fixed states; a certain scale of equipment-level power consumption data is used as a training set, and the digital filtering method is used to process the appropriate dynamic load equipment load curve, and the time-varying pulse part of the load fluctuates frequently.
  • FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • the electronic device may include:
  • Memory 401 Memory 401, processor 402, and a computer program stored on memory 401 and executable on processor 402.
  • the processor 402 executes the program, it implements the load decomposition method for industrial facilities based on mixed integer programming provided in the above embodiment.
  • electronic equipment also includes:
  • Communication interface 403 is used for communication between the memory 401 and the processor 402.
  • Memory 401 is used to store computer programs that can run on the processor 402.
  • the memory 401 may include high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
  • the bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc.
  • ISA Industry Standard Architecture
  • PCI Peripheral Component
  • EISA Extended Industry Standard Architecture
  • the bus can be divided into address bus, data bus, control bus, etc. For ease of presentation, only one thick line is used in Figure 4, but it does not mean that there is only one bus or one type of bus.
  • the memory 401, the processor 402 and the communication interface 403 are integrated on one chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through the internal interface.
  • the processor 402 may be a central processing unit (Central Processing Unit, referred to as CPU), or a specific integrated circuit (Application Specific Integrated Circuit, referred to as ASIC), or one or more processors configured to implement the embodiments of the present application. integrated circuit.
  • CPU Central Processing Unit
  • ASIC Application Specific Integrated Circuit
  • Embodiments of the present application also provide a computer-readable storage medium on which a computer program is stored.
  • the program is executed by a processor, the above load decomposition method for industrial facilities based on mixed integer programming is implemented.
  • references to the terms “one embodiment,” “some embodiments,” “an example,” “specific examples,” or “some examples” or the like means that specific features are described in connection with the embodiment or example. , structures, materials or features are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, those skilled in the art may combine and combine different embodiments or examples and features of different embodiments or examples described in this specification unless they are inconsistent with each other.
  • first and second are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Therefore, features defined as “first” and “second” may explicitly or implicitly include at least one of these features. In the description of this application, “N” means at least two, such as two, three, etc., unless otherwise clearly and specifically limited.
  • a "computer-readable medium” may be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
  • Non-exhaustive list of computer readable media include the following: electrical connections with one or N wires (electronic device), portable computer disk cartridge (magnetic device), random access memory (RAM), Read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM).
  • the computer-readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, and subsequently edited, interpreted, or otherwise suitable as necessary. process to obtain the program electronically and then store it in computer memory.
  • N steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system.
  • a suitable instruction execution system For example, if it is implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: discrete logic gate circuits with logic functions for implementing data signals; Logic circuits, application specific integrated circuits with suitable combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.
  • the program can be stored in a computer-readable storage medium.
  • the program can be stored in a computer-readable storage medium.
  • each functional unit in various embodiments of the present application can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module.
  • the above integrated modules can be implemented in the form of hardware or software function modules. If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.
  • the storage media mentioned above can be read-only memory, magnetic disks or optical disks, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Power Engineering (AREA)
  • Data Mining & Analysis (AREA)
  • Marketing (AREA)
  • Evolutionary Computation (AREA)
  • General Business, Economics & Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Quality & Reliability (AREA)
  • Evolutionary Biology (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • Computer Hardware Design (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Water Supply & Treatment (AREA)
  • Public Health (AREA)
  • Geometry (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

本申请涉及一种基于混合整数规划的针对工业设施的负荷分解方法和装置，方法包括：获取待分解工业设施的设备资产清单，将工业设备分类，一类为功率可连续调节的动态负荷设备，一类为功率在固定状态间切换的稳定负荷设备；一定规模的设备级的用电数据作为训练集，使用数字滤波的方法处理合适的动态负荷设备负荷曲线，利用负荷中波动频繁的脉冲部分的时不变性扩充其训练数据集；利用设备信息和训练数据构建用于分解的混合整数规划模型；利用一些设备的工业流程依赖等物理限制，引入模型修正项或约束；求解规划问题得到分解结果。该方法能够同时考虑动态负荷设备和稳定负荷设备的负荷特性，在工业场景下能够进行准确度较高的非侵入式负荷分解。

Description

基于混合整数规划的针对工业设施的负荷分解方法和装置
相关申请的交叉引用
本申请基于申请号为CN202210285953.1，申请日为2022年3月22日申请的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本申请作为参考。
技术领域
本申请涉及非侵入式负荷分解技术领域，特别涉及一种基于混合整数规划的针对工业设施的负荷分解方法和装置。
背景技术
随着智能电网和城市化的快速发展，海量信息的流动已成为必然趋势。电力传感器已被广泛应用于监测各类用电用户的负荷信息。这些传感器收集了大量的负载数据，为开发基于物联网的智能电网奠定了基础。另一方面，可再生能源越来越受到人们的关注。大量可再生能源的普及和燃煤电厂的逐步淘汰给电网运行带来了巨大的挑战。为了更好地促进节能，促进电网供需双方的双向通信，电力用户的设备级负荷数据反馈变得非常重要，该反馈可以通过侵入式和非侵入式两种方式实现。侵入式监控通常涉及传感器安装、经济成本、数据隐私等障碍。相比之下，非侵入式负载监测为消费者提供了成本最低、传感器安装工作量最低的解决方案，被认为是更有前景的选择。
由于不同电力用户的电力消耗规模和电气设备类型存在巨大差异，负荷分解通常分为三种类型:民用、商业和工业设置的负荷分解。尽管负荷分解在商业和工业场景中的应用具有更显著的潜在利润(一个工厂的用电量可能相当于数百个住宅的用电量)，但目前负荷分解的研究主要集中在民用设施上。与民用场景不同，工业场景存在大量的动态负荷设备，用电特征更为复杂，这给负荷分解带来了很大的困难。
在工业NILM(non-intrusive load monitoring，非侵入式电力负荷监测)方面，相关技术中首先研究了大型工业建筑中的负荷分解，通过从一个大型工业冷库收集电力数据，并将收集到的数据与典型的住宅数据进行对比分析，并且采用CO(Combinationrial Optimization，组合优化算法)和FHMM(Factorial Hidden Markov Models，多条隐马尔科夫模型)两种基准模型对工业负荷进行了分解，并提出了针对性培训和分计量的改进措施。
相关技术中还建立了一个基于深度神经网络的分解模型，取得了比经典FHMM更好的结果。
相关技术中还利用FHMM中的有功功率和无功功率数据进行工业负荷分解，在一定程度上改善了结果。目前，大多数关于工业NILM的研究只是对住宅NILM模型进行了轻微的修改，以处理工业场景，而没有设计合适的模型来利用工业数据的特征。
工业负荷具有民用负荷少有的大比例动态负荷和流水线依赖等特征，这些特征会极大地影响传统的基于优化模型和基于FHMM模型的效果。采用深度学习方法的相关研究表明，在工业场景下，深度学习方法仍能取得很好的效果，同时这些模型高度依赖于大量的训练数据，但工业数据往往十分敏感，这成为模型应用的一大障碍。
相关技术中主要存在以下两种求解技术：
(1)k-均值聚类技术：该技术可以在给定聚类中心数目k的情况下通过迭代将数据集按照距离划分为k类。
(2)混合整数规划问题求解技术：求解数学规划问题中自变量存在整数的一类问题，一般用基于分支定界的算法来解。分支定界具体的做法如下：
对最初的混整规划删除所有的整数约束，得到原规划松弛，一般这个松弛后的规划认为是可以高效求解的。如果松弛问题的解恰好满足所有整数约束的限制，那么这个解是原混整规划的最优解，运算终止。如果解没有满足所有整数限制(大多数是这种情况)，需要引入约束，将这个解排除出可行域，这是原可行域被划分为两个区域，原问题可以分为两个子问题求解，称作“分支”。这时原问题被两个整数变量更少的子问题取代了。
上述松弛问题的解，是在更大的可行域下的最优解，原问题的解一定不会比这个解更优，于是松弛问题的解被规定为原问题的一个“下界”。在分支的情况下，会不断地分出子问题，这是可能会得到满足子问题所有约束的最优解，这个解只是原问题局部可行域内的最优，不一定是全局最优，因此该解被规定为原问题的一个“上界”。在对每一次的分支子问题计算时，如果求解一个子问题落到当前所定的界之外，就可以删掉这个分支，不再进一步考虑。因此分支定界法是一个迭代算法，随着子问题的求解，不断更新上下界，当算法满足收敛条件的时候，就可以得到原问题数值上的解了。
目前一般采用成熟的商业求解器求解这类问题，这些求解器一般采用了许多算法对求解速度进行优化，如割平面法等。
发明内容
本申请提供一种基于混合整数规划的针对工业设施的负荷分解方法、装置、电子设备及存储介质，以解决现有负荷分解技术难以在工业负荷数据不充足、数据获取困难的限制下有效处理工业场景下的复杂设备负荷辨识的问题，能够同时考虑动态负荷设备和稳定负荷设备的负荷特性，在工业场景下能够进行准确度较高的非侵入式负荷分解。
本申请第一方面实施例提供一种基于混合整数规划的针对工业设施的负荷分解，包括以下步骤：
获取待分解工业设备的设备清单、待分解负荷曲线，并基于所述设备清单和所述待分解负荷曲线得到各设备的设备级负荷曲线；
根据耗电特性将所述待分解工业设备划分为动态负荷设备和稳定负荷设备，并对所述动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备的负荷曲线；
基于所述处理后的动态负荷设备负荷曲线和所述稳定负荷设备负荷曲线构建混合整数规划模型，在所述混合整数规划模型中引入工业流程约束和设备运行限制的修正后，求解得到优化变量最优取值；以及
基于所述优化问题模型对所述处理后的动态负荷设备组合信号矩阵重建出分解结果，并基于所述优化问题模型对所述稳定负荷设备转换为功率状态序列重建出分解结果。
可选地，所述对所述动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备，包括：
对所述动态负荷设备的设备级负荷曲线先利用中值滤波去除噪声尖峰，再检测所述动态负荷设备的设备级负荷曲线中的大斜率部分将其分离，最后将每个动态设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分；
生成预设时间段的元素全为0的列向量，将所述列向量连接到每个待扩充基向量的开头，并舍去所述待扩充基向量末尾同样长度的序列，得到滞后一个单位的扩充脉冲，将所述扩充脉冲叠加到所述平滑的基本部分，得到了新基向量；
基于所述新基向量和所述波动剧烈的脉冲部分得到处理后的动态负荷设备。
可选地，所述根据耗电特性将所述待分解工业设备划分为动态负荷设备和稳定负荷设备，包括：
对所述稳定负荷设备采用组合优化的方法建模：
Figure PCTCN2022135019-appb-000001
Figure PCTCN2022135019-appb-000002
Figure PCTCN2022135019-appb-000003
其中，
Figure PCTCN2022135019-appb-000004
为t时刻设备n的功率，t为时刻，T为一次优化的时间序列长度，n为设备编号，b n,k(t)为确定设备n所在的功率状态的变量，k为[1，K n]中的整数，K n为设备n的状态个数；对所述动态负荷设备采用各设备已知的设备级负荷曲线的线性组合来建模：
Figure PCTCN2022135019-appb-000005
s.t.:A n≥0；
其中，
Figure PCTCN2022135019-appb-000006
为设备n的功率时间序列，D n为设备n的信号矩阵；A n为激活系数矩阵。
可选地，所述基于所述处理后的动态负荷设备负荷曲线和所述稳定负荷设备负荷曲线构建混合整数规划模型，还包括：
将各所述动态负荷设备的信号矩阵连接为D＝[D 1,D 2,…,D n]；将所述稳定负荷设备的负荷模型表达为矩阵的形式：
Figure PCTCN2022135019-appb-000007
取和信号矩阵中基向量相同的维度，有：
Figure PCTCN2022135019-appb-000008
构建出所述混合整数规划模型：
Figure PCTCN2022135019-appb-000009
s.t.:A≥0
其中，X为用户总智能电表读出的负荷曲线，β为正则项系数，U为1构成的列向量，Q为信号分组矩阵，A为激活系数矩阵，N为稳定符合设备台数，X n为功率状态矩阵，B n为状态选择矩阵，m是一次优化的截取的时间序列长度。
可选地，其特征在于，所述在所述混合整数规划模型中引入工业流程约束和设备运行限制的修正，包括：
流程约束为：
Figure PCTCN2022135019-appb-000010
对所述稳定负荷设备限制最长持续时间和限制最短持续时间，其中，所述最长持续时间W n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000011
Figure PCTCN2022135019-appb-000012
所述最短持续时间w n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000013
Figure PCTCN2022135019-appb-000014
其中，γ为正则项系数，R为流程限制矩阵。
本申请第二方面实施例提供一种基于混合整数规划的针对工业设施的负荷分解装置，包括：
获取模块，用于获取待分解工业设备的设备清单、待分解负荷曲线，并基于所述设备清单和所述待分解负荷曲线得到各设备的设备级负荷曲线；
处理模块，用于根据耗电特性将所述待分解工业设备划分为动态负荷设备和稳定负荷设备，并对所述动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备的负荷曲线；
第一计算模块，用于基于所述处理后的动态负荷设备负荷曲线和所述稳定负荷设备负荷曲线构建混合整数规划模型，在所述混合整数规划模型中引入工业流程约束和设备运行限制的修正后，求解得到优化变量最优取值；以及
第二计算模块，用于基于所述优化问题模型对所述处理后的动态负荷设备组合信号矩阵重建出分解结果，并基于所述优化问题模型对所述稳定负荷设备转换为功率状态序列重建出分解结果。
可选地，所述处理模块，具体用于：
对所述动态负荷设备的设备级负荷曲线先利用中值滤波去除噪声尖峰，再检测所述动态负荷设备的设备级负荷曲线中的大斜率部分将其分离，最后将每个动态设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分；
生成预设时间段的元素全为0的列向量，将所述列向量连接到每个待扩充基向量的开头，并舍去所述待扩充基向量末尾同样长度的序列，得到滞后一个单位的扩充脉冲，将所述扩充脉冲叠加到所述平滑的基本部分，得到了新基向量；
基于所述新基向量和所述波动剧烈的脉冲部分得到处理后的动态负荷设备。
可选地，所述处理模块用于：
对所述稳定负荷设备采用组合优化的方法建模：
Figure PCTCN2022135019-appb-000015
Figure PCTCN2022135019-appb-000016
Figure PCTCN2022135019-appb-000017
其中，
Figure PCTCN2022135019-appb-000018
为t时刻设备n的功率，t为时刻，T为一次优化的时间序列长度，n为设备编号，b n,k(t)为确定设备n所在的功率状态的变量，k为[1，K n]中的整数，K n为设备n的状 态个数；
对所述动态负荷设备采用各设备已知的设备级负荷曲线的线性组合来建模：
Figure PCTCN2022135019-appb-000019
s.t.:A n≥0；
其中，
Figure PCTCN2022135019-appb-000020
为设备n的功率时间序列，D n为设备n的信号矩阵；A n为激活系数矩阵。
可选地，所述第一计算模块，还用于：
将各所述动态负荷设备的信号矩阵连接为D＝[D 1,D 2,…,D n]；将所述稳定负荷设备的负荷模型表达为矩阵的形式：
Figure PCTCN2022135019-appb-000021
取和信号矩阵中基向量相同的维度，有：
Figure PCTCN2022135019-appb-000022
构建出所述混合整数规划模型：
Figure PCTCN2022135019-appb-000023
s.t.:A≥0
其中，X为用户总智能电表读出的负荷曲线，β为正则项系数，U为1构成的列向量，Q为信号分组矩阵，A为激活系数矩阵，N为稳定符合设备台数，X n为功率状态矩阵，B n为状态选择矩阵，m是一次优化的截取的时间序列长度。
可选地，所述第一计算模块，还用于：
流程约束为：
Figure PCTCN2022135019-appb-000024
对所述稳定负荷设备限制最长持续时间和限制最短持续时间，其中，所述最长持续时间W n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000025
Figure PCTCN2022135019-appb-000026
所述最短持续时间w n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000027
Figure PCTCN2022135019-appb-000028
其中，γ为正则项系数，R为流程限制矩阵。
本申请第三方面实施例提供一种电子设备，包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器执行所述程序，以实现如上述实施例所述的基于混合整数规划的针对工业设施的负荷分解方法。
本申请第四方面实施例提供一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行，以用于实现如上述的基于混合整数规划的针对工业设施的负荷分解方法。
由此，通过获取待分解工业设施的设备资产清单，将工业设备分类，一类为功率可连续调节的动态负荷设备，一类为功率在固定状态间切换的稳定负荷设备；一定规模的设备级的用电数据作为训练集，使用数字滤波的方法处理合适的动态负荷设备负荷曲线，利用负荷中波动频繁的脉冲部分的时不变性扩充其训练数据集；利用设备信息和训练数据构建用于分解的混合整数规划模型；利用一些设备的工业流程依赖等物理限制，引入模型修正项或约束；求解规划问题得到分解结果。由此，解决了现有负荷分解技术难以在工业负荷数据不充足、数据获取困难的限制下有效处理工业场景下的复杂设备负荷辨识的问题，能够同时考虑动态负荷设备和稳定负荷设备的负荷特性，在工业场景下能够进行准确度较高的非侵入式负荷分解。
本申请附加的方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本申请的实践了解到。
附图说明
本申请上述的和/或附加的方面和优点从下面结合附图对实施例的描述中将变得明显和容易理解，其中：
图1为根据本申请实施例提供的基于混合整数规划的针对工业设施的负荷分解方法的流程图；
图2为根据本申请一个实施例提供的基于混合整数规划的针对工业设施的负荷分解方法的流程图；
图3为根据本申请实施例提供的基于混合整数规划的针对工业设施的负荷分解装置的方框示意图；
图4为根据本申请实施例的电子设备的示意图。
具体实施方式
下面详细描述本申请的实施例，所述实施例的示例在附图中示出，其中自始至终相同 或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，旨在用于解释本申请，而不能理解为对本申请的限制。
下面参考附图描述本申请实施例的基于混合整数规划的针对工业设施的负荷分解方法、装置、电子设备及存储介质。针对上述背景技术中心提到的现有负荷分解技术难以在工业负荷数据不充足、数据获取困难的限制下有效处理工业场景下的复杂设备负荷辨识的问题，本申请提供了一种基于混合整数规划的针对工业设施的负荷分解方法，在该方法中，通过获取待分解工业设施的设备资产清单，将工业设备分类，一类为功率可连续调节的动态负荷设备，一类为功率在固定状态间切换的稳定负荷设备；以及一定规模的设备级的用电数据作为训练集，使用数字滤波的方法处理合适的动态负荷设备负荷曲线，利用负荷中波动频繁的脉冲部分的时不变性扩充其训练数据集；利用设备信息和训练数据构建用于分解的混合整数规划模型；利用一些设备的工业流程依赖等物理限制，引入模型修正项或约束；求解规划问题得到分解结果。由此，解决了现有负荷分解技术难以在工业负荷数据不充足、数据获取困难的限制下有效处理工业场景下的复杂设备负荷辨识的问题，能够同时考虑动态负荷设备和稳定负荷设备的负荷特性，在工业场景下能够进行准确度较高的非侵入式负荷分解。
具体而言，图1为本申请实施例所提供的一种基于混合整数规划的针对工业设施的负荷分解方法的流程示意图。
如图1所示，该基于混合整数规划的针对工业设施的负荷分解方法包括以下步骤：
在步骤S101中，获取待分解工业设备的设备清单、待分解负荷曲线，并基于设备清单和待分解负荷曲线得到各设备的设备级负荷曲线。
其中，定义用户低压回路智能电表在一天中记录的功率读数序列为“总负荷曲线”，并记录为一个列向量；定义对用电设备单独安装传感器，采集的一天的功率读数序列为“设备级负荷曲线”，记录为同样维度的列向量。
具体地，获取待分解工业设施的设备资产清单；从待分解工业设施智能电表获取待分解负荷曲线，从设备产商或者设备负荷数据库中获取各设备的设备级负荷曲线。
在步骤S102中，根据耗电特性将待分解工业设备划分为动态负荷设备和稳定负荷设备，并对动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备的负荷曲线。
可选地，在一些实施例中，根据耗电特性将待分解工业设备划分为动态负荷设备和稳定负荷设备，包括：对稳定负荷设备采用组合优化的方法建模：
Figure PCTCN2022135019-appb-000029
Figure PCTCN2022135019-appb-000030
Figure PCTCN2022135019-appb-000031
其中，
Figure PCTCN2022135019-appb-000032
为t时刻设备n的功率，t为时刻，T为一次优化的时间序列长度，n为设备编号，设备n的各个状态的功率大小记为
Figure PCTCN2022135019-appb-000033
各状态的功率值采用k-均值聚类技术确定，b n,k(t)为确定设备n所在的功率状态的变量，k为[1，K n]中的整数，K n为设备n的状态个数；
对动态负荷设备采用各设备已知的设备级负荷曲线的线性组合来建模：
Figure PCTCN2022135019-appb-000034
s.t.:A n≥0；
其中，
Figure PCTCN2022135019-appb-000035
为设备n的功率时间序列，D n为设备n的信号矩阵，其每个列向量为该设备已知的设备级负荷曲线，称为一个基向量；A n为激活系数矩阵，即对基向量进行线性组合的系数，一般要求A的元素非负。
可选地，在一些实施例中，对动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备，包括：对动态负荷设备的设备级负荷曲线先利用中值滤波去除噪声尖峰，再检测动态负荷设备的设备级负荷曲线中的大斜率部分将其分离，最后将每个动态设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分；生成预设时间段的元素全为0的列向量，将列向量连接到每个待扩充基向量的开头，并舍去待扩充基向量末尾同样长度的序列，得到滞后一个单位的扩充脉冲，将扩充脉冲叠加到平滑的基本部分，得到了新基向量；基于新基向量和波动剧烈的脉冲部分得到处理后的动态负荷设备。
具体地，根据设备工作时的耗电特性将工业设备分类，将设备分为动态负荷设备和稳定负荷设备；动态负荷设备一般包含可调速电机，其负荷曲线在连续范围内变化；稳定负荷设备负荷曲线在一系列固定的功率状态间切换；这两类设备采用不同的方式建模。对动态负荷设备的设备级负荷曲线，使用数字滤波的方法处理合适设备对应的曲线向量，利用负荷中波动频繁的脉冲部分的时不变性构造新的符合实际的向量，从而扩充训练集；对动态负荷设备的设备级负荷曲线，首先进行数字滤波；滤波时先利用中值滤波去除噪声尖峰，然后检测曲线中的大斜率部分将其分离，另外加入限位器辅助脉冲的识别，最终将每个设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分。利用时不变性扩充脉冲部分；取一个固定的时间长度，生成对应该时间长度的元素全为0的列向量，将该列向量连接到每个待扩充基向量的开头，并舍去该基向量末尾同样长度的序列，就得到了滞后一个单位的扩充脉冲；将扩充脉冲叠加到基本部分，就得到了该设备一个新的基向量。
在步骤S103中，基于处理后的动态负荷设备负荷曲线和稳定负荷设备负荷曲线构建混 合整数规划模型，在混合整数规划模型中引入工业流程约束和设备运行限制的修正后，求解得到优化变量最优取值。
可选地，在一些实施例中，基于处理后的动态负荷设备负荷曲线和稳定负荷设备负荷曲线构建混合整数规划模型，还包括：将各动态负荷设备的信号矩阵连接为D＝[D 1,D 2,…,D n]；将稳定负荷设备的负荷模型表达为矩阵的形式：
Figure PCTCN2022135019-appb-000036
取和信号矩阵中基向量相同的维度，有：
Figure PCTCN2022135019-appb-000037
构建出混合整数规划模型：
Figure PCTCN2022135019-appb-000038
s.t.:A≥0
其中，X为用户总智能电表读出的负荷曲线，为待分解负荷，第二项为基向量分组的正则项，β为正则项系数，U为1构成的列向量，Q为信号分组矩阵，其中，Q的对角分块上为1，维度对应于所分组的基向量的个数，其他区域元素为0，A为激活系数矩阵，N为稳定符合设备台数，X n为功率状态矩阵，B n为状态选择矩阵，m是一次优化的截取的时间序列长度。
可选地，在一些实施例中，在混合整数规划模型中引入工业流程约束和设备运行限制的修正，包括：流程约束为在优化目标函数(混合整数规划模型)里加入额外惩罚项：
Figure PCTCN2022135019-appb-000039
对稳定负荷设备限制最长持续时间和限制最短持续时间，其中，最长持续时间W n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000040
Figure PCTCN2022135019-appb-000041
最短持续时间w n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000042
Figure PCTCN2022135019-appb-000043
其中，γ为正则项系数，R为流程限制矩阵，其中，R由一系列单位矩阵连接而成，两两存在流程关系的设备，对应的分块分别为正的单位矩阵和负的单位矩阵。
在步骤S104中，基于优化问题模型对处理后的动态负荷设备组合信号矩阵重建出分解结果，并基于优化问题模型对稳定负荷设备转换为功率状态序列重建出分解结果。
具体地，采用混合整数规划求解技术，求解上述优化问题得到A,B n，通过负荷模型DA重建动态负荷设备负荷，通过B nX n重建文风负荷设备负荷，从而得到分解结果。
由此，本申请综合了经典的组合优化模型在处理稳定负荷设备时的优势和矩阵分解模型在处理动态负荷设备时的优势，在一个优化规划问题中统一求解，相比于其他基于概率模型和基于优化模型的算法，更加适合不同负荷类型混杂的复杂工业场景。另一方面，本申请同时考虑了数据依赖问题，相较于应用广泛的深度学习模型，本申请能够在使用更少的训练数据的情况下，达到与一般深度学习模型同等水平的分解效果。本申请还能够在工业实际应用场景下，充分利用已有数据，获得更加准确的负荷分解结果，具有重要的意义和良好的应用前景。
为使本领域技术人员进一步了解本基于混合整数规划的针对工业设施的负荷分解方法，下面结合具体实施例进行阐述。
图2是本基于混合整数规划的针对工业设施的负荷分解方法的流程图。
S201，获取设备清单；获取待分解负荷曲线和设备级负荷曲线。
S202，将设备分类为动态负荷设备，用D nA n表示，和稳定负荷设备，负荷用
Figure PCTCN2022135019-appb-000044
表示。
S203，对动态负荷设备的设备级负荷曲线数字滤波，利用时不变性扩充脉冲部分，重建后实现基向量的扩充。
S204，考虑不同类型的设备，建立混合整数规划模型。
S205，考虑工业生产的物理限制，在模型中引入针对工业流程约束和设备运行限制的修正。
S206，利用混合规划求解技术求解优化问题，重建负荷分解结果。
根据本申请实施例提出的基于混合整数规划的针对工业设施的负荷分解方法，通过获取待分解工业设施的设备资产清单，将工业设备分类，一类为功率可连续调节的动态负荷设备，一类为功率在固定状态间切换的稳定负荷设备；一定规模的设备级的用电数据作为训练集，使用数字滤波的方法处理合适的动态负荷设备负荷曲线，利用负荷中波动频繁的脉冲部分的时不变性扩充其训练数据集；利用设备信息和训练数据构建用于分解的混合整数规划模型；利用一些设备的工业流程依赖等物理限制，引入模型修正项或约束；求解规 划问题得到分解结果。由此，解决了现有负荷分解技术难以在工业负荷数据不充足、数据获取困难的限制下有效处理工业场景下的复杂设备负荷辨识的问题，能够同时考虑动态负荷设备和稳定负荷设备的负荷特性，在工业场景下能够进行准确度较高的非侵入式负荷分解。
其次参照附图描述根据本申请实施例提出的基于混合整数规划的针对工业设施的负荷分解装置。
图3是本申请实施例的基于混合整数规划的针对工业设施的负荷分解装置的方框示意图。
如图3所示，该基于混合整数规划的针对工业设施的负荷分解装置10包括：获取模块100、处理模块200、第一计算模块300和二计算模块400。
其中，获取模块100，用于获取待分解工业设备的设备清单、待分解负荷曲线，并基于设备清单和待分解负荷曲线得到各设备的设备级负荷曲线；
处理模块200，用于根据耗电特性将待分解工业设备划分为动态负荷设备和稳定负荷设备，并对动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备的负荷曲线；
第一计算模块300，用于基于处理后的动态负荷设备负荷曲线和稳定负荷设备负荷曲线构建混合整数规划模型，在混合整数规划模型中引入工业流程约束和设备运行限制的修正后，求解得到优化变量最优取值；以及
第二计算模块400，用于基于优化问题模型对处理后的动态负荷设备组合信号矩阵重建出分解结果，并基于优化问题模型对稳定负荷设备转换为功率状态序列重建出分解结果。
可选地，在一些实施例中，处理模块200，具体用于：
对动态负荷设备的设备级负荷曲线先利用中值滤波去除噪声尖峰，再检测动态负荷设备的设备级负荷曲线中的大斜率部分将其分离，最后将每个动态设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分；
生成预设时间段的元素全为0的列向量，将列向量连接到每个待扩充基向量的开头，并舍去待扩充基向量末尾同样长度的序列，得到滞后一个单位的扩充脉冲，将扩充脉冲叠加到平滑的基本部分，得到了新基向量；
基于新基向量和波动剧烈的脉冲部分得到处理后的动态负荷设备。
可选地，处理模块200，还用于：
对稳定负荷设备采用组合优化的方法建模：
Figure PCTCN2022135019-appb-000045
Figure PCTCN2022135019-appb-000046
Figure PCTCN2022135019-appb-000047
其中，
Figure PCTCN2022135019-appb-000048
为t时刻设备n的功率，t为时刻，T为一次优化的时间序列长度，n为设备编号，b n,k(t)为确定设备n所在的功率状态的变量，k为[1，K n]中的整数，K n为设备n的状态个数；
对动态负荷设备采用各设备已知的设备级负荷曲线的线性组合来建模：
Figure PCTCN2022135019-appb-000049
s.t.:A n≥0；
其中，
Figure PCTCN2022135019-appb-000050
为设备n的功率时间序列，D n为设备n的信号矩阵；A n为激活系数矩阵。
可选地，第一计算模块300，还用于：
将各动态负荷设备的信号矩阵连接为D＝[D 1,D 2,…,D n]；将稳定负荷设备的负荷模型表达为矩阵的形式：
Figure PCTCN2022135019-appb-000051
取和信号矩阵中基向量相同的维度，有：
Figure PCTCN2022135019-appb-000052
构建出混合整数规划模型：
Figure PCTCN2022135019-appb-000053
s.t.:A≥0
其中，X为用户总智能电表读出的负荷曲线，β为正则项系数，U为1构成的列向量，Q为信号分组矩阵，A为激活系数矩阵，N为稳定符合设备台数，X n为功率状态矩阵，B n为状态选择矩阵，m是一次优化的截取的时间序列长度。
可选地，在一些实施例中，第一计算模块300，还用于：
流程约束为：
Figure PCTCN2022135019-appb-000054
对稳定负荷设备限制最长持续时间和限制最短持续时间，其中，最长持续时间W n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000055
Figure PCTCN2022135019-appb-000056
最短持续时间w n,k采用一组线性不等式来约束：
Figure PCTCN2022135019-appb-000057
Figure PCTCN2022135019-appb-000058
其中，γ为正则项系数，R为流程限制矩阵。
需要说明的是，前述对基于混合整数规划的针对工业设施的负荷分解方法实施例的解释说明也适用于该实施例的基于混合整数规划的针对工业设施的负荷分解装置，此处不再赘述。
根据本申请实施例提出的基于混合整数规划的针对工业设施的负荷分解装置，通过获取待分解工业设施的设备资产清单，将工业设备分类，一类为功率可连续调节的动态负荷设备，一类为功率在固定状态间切换的稳定负荷设备；一定规模的设备级的用电数据作为训练集，使用数字滤波的方法处理合适的动态负荷设备负荷曲线，利用负荷中波动频繁的脉冲部分的时不变性扩充其训练数据集；利用设备信息和训练数据构建用于分解的混合整数规划模型；利用一些设备的工业流程依赖等物理限制，引入模型修正项或约束；求解规划问题得到分解结果。由此，解决了现有负荷分解技术难以在工业负荷数据不充足、数据获取困难的限制下有效处理工业场景下的复杂设备负荷辨识的问题，能够同时考虑动态负荷设备和稳定负荷设备的负荷特性，在工业场景下能够进行准确度较高的非侵入式负荷分解。
图4为本申请实施例提供的电子设备的结构示意图。该电子设备可以包括：
存储器401、处理器402及存储在存储器401上并可在处理器402上运行的计算机程序。
处理器402执行程序时实现上述实施例中提供的基于混合整数规划的针对工业设施的负荷分解方法。
进一步地，电子设备还包括：
通信接口403，用于存储器401和处理器402之间的通信。
存储器401，用于存放可在处理器402上运行的计算机程序。
存储器401可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。
如果存储器401、处理器402和通信接口403独立实现，则通信接口403、存储器401 和处理器402可以通过总线相互连接并完成相互间的通信。总线可以是工业标准体系结构(Industry Standard Architecture，简称为ISA)总线、外部设备互连(Peripheral Component，简称为PCI)总线或扩展工业标准体系结构(Extended Industry Standard Architecture，简称为EISA)总线等。总线可以分为地址总线、数据总线、控制总线等。为便于表示，图4中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
可选的，在具体实现上，如果存储器401、处理器402及通信接口403，集成在一块芯片上实现，则存储器401、处理器402及通信接口403可以通过内部接口完成相互间的通信。
处理器402可能是一个中央处理器(Central Processing Unit，简称为CPU)，或者是特定集成电路(Application Specific Integrated Circuit，简称为ASIC)，或者是被配置成实施本申请实施例的一个或多个集成电路。
本申请实施例还提供一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现如上的基于混合整数规划的针对工业设施的负荷分解方法。
在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本申请的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不必须针对的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任一个或N个实施例或示例中以合适的方式结合。此外，在不相互矛盾的情况下，本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。
此外，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。在本申请的描述中，“N个”的含义是至少两个，例如两个，三个等，除非另有明确具体的限定。
流程图中或在此以其他方式描述的任何过程或方法描述可以被理解为，表示包括一个或更N个用于实现定制逻辑功能或过程的步骤的可执行指令的代码的模块、片段或部分，并且本申请的优选实施方式的范围包括另外的实现，其中可以不按所示出或讨论的顺序，包括根据所涉及的功能按基本同时的方式或按相反的顺序，来执行功能，这应被本申请的实施例所属技术领域的技术人员所理解。
在流程图中表示或在此以其他方式描述的逻辑和/或步骤，例如，可以被认为是用于实 现逻辑功能的可执行指令的定序列表，可以具体实现在任何计算机可读介质中，以供指令执行系统、装置或设备(如基于计算机的系统、包括处理器的系统或其他可以从指令执行系统、装置或设备取指令并执行指令的系统)使用，或结合这些指令执行系统、装置或设备而使用。就本说明书而言，"计算机可读介质"可以是任何可以包含、存储、通信、传播或传输程序以供指令执行系统、装置或设备或结合这些指令执行系统、装置或设备而使用的装置。计算机可读介质的更具体的示例(非穷尽性列表)包括以下：具有一个或N个布线的电连接部(电子装置)，便携式计算机盘盒(磁装置)，随机存取存储器(RAM)，只读存储器(ROM)，可擦除可编辑只读存储器(EPROM或闪速存储器)，光纤装置，以及便携式光盘只读存储器(CDROM)。另外，计算机可读介质甚至可以是可在其上打印所述程序的纸或其他合适的介质，因为可以例如通过对纸或其他介质进行光学扫描，接着进行编辑、解译或必要时以其他合适方式进行处理来以电子方式获得所述程序，然后将其存储在计算机存储器中。
应当理解，本申请的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式中，N个步骤或方法可以用存储在存储器中且由合适的指令执行系统执行的软件或固件来实现。如，如果用硬件来实现和在另一实施方式中一样，可用本领域公知的下列技术中的任一项或他们的组合来实现：具有用于对数据信号实现逻辑功能的逻辑门电路的离散逻辑电路，具有合适的组合逻辑门电路的专用集成电路，可编程门阵列(PGA)，现场可编程门阵列(FPGA)等。
本技术领域的普通技术人员可以理解实现上述实施例方法携带的全部或部分步骤是可以通过程序来指令相关的硬件完成，所述的程序可以存储于一种计算机可读存储介质中，该程序在执行时，包括方法实施例的步骤之一或其组合。
此外，在本申请各个实施例中的各功能单元可以集成在一个处理模块中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个模块中。上述集成的模块既可以采用硬件的形式实现，也可以采用软件功能模块的形式实现。所述集成的模块如果以软件功能模块的形式实现并作为独立的产品销售或使用时，也可以存储在一个计算机可读取存储介质中。
上述提到的存储介质可以是只读存储器，磁盘或光盘等。尽管上面已经示出和描述了本申请的实施例，可以理解的是，上述实施例是示例性的，不能理解为对本申请的限制，本领域的普通技术人员在本申请的范围内可以对上述实施例进行变化、修改、替换和变型。

Claims (12)

  1. 一种基于混合整数规划的针对工业设施的负荷分解方法，其特征在于，包括以下步骤：
    获取待分解工业设备的设备清单、待分解负荷曲线，并基于所述设备清单和所述待分解负荷曲线得到各设备的设备级负荷曲线；
    根据耗电特性将所述待分解工业设备划分为动态负荷设备和稳定负荷设备，并对所述动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备的负荷曲线；
    基于所述处理后的动态负荷设备负荷曲线和所述稳定负荷设备负荷曲线构建混合整数规划模型，在所述混合整数规划模型中引入工业流程约束和设备运行限制的修正后，求解得到优化变量最优取值；以及
    基于所述优化问题模型对所述处理后的动态负荷设备组合信号矩阵重建出分解结果，并基于所述优化问题模型对所述稳定负荷设备转换为功率状态序列重建出分解结果。
  2. 根据权利要求1所述的方法，其特征在于，所述对所述动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备，包括：
    对所述动态负荷设备的设备级负荷曲线先利用中值滤波去除噪声尖峰，再检测所述动态负荷设备的设备级负荷曲线中的大斜率部分将其分离，最后将每个动态设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分；
    生成预设时间段的元素全为0的列向量，将所述列向量连接到每个待扩充基向量的开头，并舍去所述待扩充基向量末尾同样长度的序列，得到滞后一个单位的扩充脉冲，将所述扩充脉冲叠加到所述平滑的基本部分，得到了新基向量；
    基于所述新基向量和所述波动剧烈的脉冲部分得到处理后的动态负荷设备。
  3. 根据权利要求2所述的方法，其特征在于，所述根据耗电特性将所述待分解工业设备划分为动态负荷设备和稳定负荷设备，包括：
    对所述稳定负荷设备采用组合优化的方法建模：
    Figure PCTCN2022135019-appb-100001
    Figure PCTCN2022135019-appb-100002
    Figure PCTCN2022135019-appb-100003
    其中，
    Figure PCTCN2022135019-appb-100004
    为t时刻设备n的功率，t为时刻，T为一次优化的时间序列长度，n为设备编号，b n,k(t)为确定设备n所在的功率状态的变量，k为[1，K n]中的整数，K n为设备n的状态个数；
    对所述动态负荷设备采用各设备已知的设备级负荷曲线的线性组合来建模：
    Figure PCTCN2022135019-appb-100005
    其中，
    Figure PCTCN2022135019-appb-100006
    为设备n的功率时间序列，D n为设备n的信号矩阵；A n为激活系数矩阵。
  4. 根据权利要求3所述的方法，其特征在于，所述基于所述处理后的动态负荷设备负荷曲线和所述稳定负荷设备负荷曲线构建混合整数规划模型，还包括：
    将各所述动态负荷设备的信号矩阵连接为D＝[D 1,D 2,…,D n]；将所述稳定负荷设备的负荷模型表达为矩阵的形式：
    Figure PCTCN2022135019-appb-100007
    取和信号矩阵中基向量相同的维度，有：
    Figure PCTCN2022135019-appb-100008
    构建出所述混合整数规划模型：
    Figure PCTCN2022135019-appb-100009
    其中，X为用户总智能电表读出的负荷曲线，β为正则项系数，U为1构成的列向量，Q为信号分组矩阵，A为激活系数矩阵，N为稳定符合设备台数，X n为功率状态矩阵，B n为状态选择矩阵，m是一次优化的截取的时间序列长度。
  5. 根据权利要求1所述的方法，其特征在于，所述在所述混合整数规划模型中引入工业流程约束和设备运行限制的修正，包括：
    流程约束为：
    Figure PCTCN2022135019-appb-100010
    对所述稳定负荷设备限制最长持续时间和限制最短持续时间，其中，所述最长持续时间W n,k采用一组线性不等式来约束：
    Figure PCTCN2022135019-appb-100011
    Figure PCTCN2022135019-appb-100012
    所述最短持续时间w n,k采用一组线性不等式来约束：
    Figure PCTCN2022135019-appb-100013
    其中，γ为正则项系数，R为流程限制矩阵。
  6. 一种基于混合整数规划的针对工业设施的负荷分解装置，其特征在于，包括：
    获取模块，用于获取待分解工业设备的设备清单、待分解负荷曲线，并基于所述设备清单和所述待分解负荷曲线得到各设备的设备级负荷曲线；
    处理模块，用于根据耗电特性将所述待分解工业设备划分为动态负荷设备和稳定负荷设备，并对所述动态负荷设备的设备级负荷曲线进行数字滤波，并利用时不变性扩充负荷中波动频繁的脉冲部分，得到处理后的动态负荷设备的负荷曲线；
    第一计算模块，用于基于所述处理后的动态负荷设备负荷曲线和所述稳定负荷设备负荷曲线构建混合整数规划模型，在所述混合整数规划模型中引入工业流程约束和设备运行限制的修正后，求解得到优化变量最优取值；以及
    第二计算模块，用于基于所述优化问题模型对所述处理后的动态负荷设备组合信号矩阵重建出分解结果，并基于所述优化问题模型对所述稳定负荷设备转换为功率状态序列重建出分解结果。
  7. 根据权利要求6所述的装置，其特征在于，所述处理模块，具体用于：
    对所述动态负荷设备的设备级负荷曲线先利用中值滤波去除噪声尖峰，再检测所述动态负荷设备的设备级负荷曲线中的大斜率部分将其分离，最后将每个动态设备的负荷分解为平滑的基本部分和波动剧烈的脉冲部分；
    生成预设时间段的元素全为0的列向量，将所述列向量连接到每个待扩充基向量的开头，并舍去所述待扩充基向量末尾同样长度的序列，得到滞后一个单位的扩充脉冲，将所述扩充脉冲叠加到所述平滑的基本部分，得到了新基向量；
    基于所述新基向量和所述波动剧烈的脉冲部分得到处理后的动态负荷设备。
  8. 根据权利要求7所述的装置，其特征在于，所述处理模块用于：
    对所述稳定负荷设备采用组合优化的方法建模：
    Figure PCTCN2022135019-appb-100014
    Figure PCTCN2022135019-appb-100015
    Figure PCTCN2022135019-appb-100016
    其中，
    Figure PCTCN2022135019-appb-100017
    为t时刻设备n的功率，t为时刻，T为一次优化的时间序列长度，n为设备编号，b n,k(t)为确定设备n所在的功率状态的变量，k为[1，K n]中的整数，K n为设备n的状态个数；
    对所述动态负荷设备采用各设备已知的设备级负荷曲线的线性组合来建模：
    Figure PCTCN2022135019-appb-100018
    其中，
    Figure PCTCN2022135019-appb-100019
    为设备n的功率时间序列，D n为设备n的信号矩阵；A n为激活系数矩阵。
  9. 根据权利要求8所述的装置，其特征在于，所述第一计算模块，还用于：
    将各所述动态负荷设备的信号矩阵连接为D＝[D 1,D 2,…,D n]；将所述稳定负荷设备的负荷模型表达为矩阵的形式：
    Figure PCTCN2022135019-appb-100020
    取和信号矩阵中基向量相同的维度，有：
    Figure PCTCN2022135019-appb-100021
    构建出所述混合整数规划模型：
    Figure PCTCN2022135019-appb-100022
    其中，X为用户总智能电表读出的负荷曲线，β为正则项系数，U为1构成的列向量，Q为信号分组矩阵，A为激活系数矩阵，N为稳定符合设备台数，X n为功率状态矩阵，B n为状态选择矩阵，m是一次优化的截取的时间序列长度。
  10. 根据权利要求6所述的装置，其特征在于，所述第一计算模块，用于：
    流程约束为：
    Figure PCTCN2022135019-appb-100023
    对所述稳定负荷设备限制最长持续时间和限制最短持续时间，其中，所述最长持续时间W n,k采用一组线性不等式来约束：
    Figure PCTCN2022135019-appb-100024
    所述最短持续时间w n,k采用一组线性不等式来约束：
    Figure PCTCN2022135019-appb-100025
    其中，γ为正则项系数，R为流程限制矩阵。
  11. 一种电子设备，其特征在于，包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器执行所述程序，以实现如权利要求1-5任一项所述的基于混合整数规划的针对工业设施的负荷分解方法。
  12. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行，以用于实现如权利要求1-5任一项所述的基于混合整数规划的针对工业设施的负荷分解方法。
PCT/CN2022/135019 2022-03-22 2022-11-29 基于混合整数规划的针对工业设施的负荷分解方法和装置 Ceased WO2023179076A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/759,982 US20240364107A1 (en) 2022-03-22 2024-06-30 Mixed integer programming-based load disaggregation method and apparatus for industrial facility

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210285953.1A CN114595591B (zh) 2022-03-22 2022-03-22 基于混合整数规划的针对工业设施的负荷分解方法和装置
CN202210285953.1 2022-03-22

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/759,982 Continuation US20240364107A1 (en) 2022-03-22 2024-06-30 Mixed integer programming-based load disaggregation method and apparatus for industrial facility

Publications (1)

Publication Number Publication Date
WO2023179076A1 true WO2023179076A1 (zh) 2023-09-28

Family

ID=81819726

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/135019 Ceased WO2023179076A1 (zh) 2022-03-22 2022-11-29 基于混合整数规划的针对工业设施的负荷分解方法和装置

Country Status (3)

Country Link
US (1) US20240364107A1 (zh)
CN (1) CN114595591B (zh)
WO (1) WO2023179076A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN119645446A (zh) * 2024-10-18 2025-03-18 重庆赛力斯凤凰智创科技有限公司 电子控制单元刷写方法和系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114595591B (zh) * 2022-03-22 2025-05-30 清华大学 基于混合整数规划的针对工业设施的负荷分解方法和装置
CN115933519B (zh) * 2022-12-21 2025-11-11 北京发那科机电有限公司 一种刀具监控方法、装置、电子设备及存储介质
CN119249076B (zh) * 2024-12-09 2025-06-10 航天亮丽电气有限责任公司 一种用电信息采集终端的数据采集方法
CN120408131A (zh) * 2025-03-21 2025-08-01 江西冠英智能科技股份有限公司 设备负载数据处理方法及电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112348096A (zh) * 2020-11-11 2021-02-09 合肥工业大学 非侵入式负荷分解方法及系统
CN113177299A (zh) * 2021-04-12 2021-07-27 航天科工深圳（集团）有限公司 非侵入式电力负荷分解方法、装置、设备及介质
CN113408210A (zh) * 2021-07-12 2021-09-17 内蒙古电力（集团）有限责任公司乌兰察布电业局 基于深度学习的非侵入负荷分解方法、系统、介质和设备
CN114595591A (zh) * 2022-03-22 2022-06-07 清华大学 基于混合整数规划的针对工业设施的负荷分解方法和装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105226643B (zh) * 2015-09-23 2017-11-28 清华大学 安全约束下电力系统运行模拟模型快速生成与求解方法
US11262386B2 (en) * 2015-10-16 2022-03-01 Massachusetts Institute Of Technology Non-intrusive monitoring
CN107730003B (zh) * 2017-10-23 2020-09-08 华中科技大学 一种支持多电器类型高精度的nilm实现方法
CN112435142B (zh) * 2020-12-16 2022-05-20 北京航空航天大学 一种用电负荷识别方法及其负荷用电设施知识库构建方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112348096A (zh) * 2020-11-11 2021-02-09 合肥工业大学 非侵入式负荷分解方法及系统
CN113177299A (zh) * 2021-04-12 2021-07-27 航天科工深圳（集团）有限公司 非侵入式电力负荷分解方法、装置、设备及介质
CN113408210A (zh) * 2021-07-12 2021-09-17 内蒙古电力（集团）有限责任公司乌兰察布电业局 基于深度学习的非侵入负荷分解方法、系统、介质和设备
CN114595591A (zh) * 2022-03-22 2022-06-07 清华大学 基于混合整数规划的针对工业设施的负荷分解方法和装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN119645446A (zh) * 2024-10-18 2025-03-18 重庆赛力斯凤凰智创科技有限公司 电子控制单元刷写方法和系统

Also Published As

Publication number Publication date
CN114595591A (zh) 2022-06-07
US20240364107A1 (en) 2024-10-31
CN114595591B (zh) 2025-05-30

Similar Documents

Publication Publication Date Title
Fan et al. A review on data preprocessing techniques toward efficient and reliable knowledge discovery from building operational data
Culaba et al. Machine learning‐based energy consumption clustering and forecasting for mixed‐use buildings
CN112529282A (zh) 基于时空图卷积神经网络的风电场集群短期功率预测方法
CN114595591B (zh) 基于混合整数规划的针对工业设施的负荷分解方法和装置
CN112381313B (zh) 充电桩地址的确定方法和装置
Krishnan et al. Efficient database generation for decision tree based power system security assessment
CN114819377A (zh) 分散式风电功率预测方法、系统、装置及存储介质
Yao et al. Electric supply and demand forecasting using seasonal grey model based on PSO-SVR
Zhang et al. Load prediction based on hybrid model of VMD‐mRMR‐BPNN‐LSSVM
Mohamed et al. A review on big data management and decision-making in smart grid
Tie et al. Research on the application of an improved Autoformer model integrating CNN-attention-BiGRU in short-term power load forecasting
CN120146291A (zh) 一种基于时空图神经网络的光伏发电能力预测方法、系统及介质
CN113807568B (zh) 一种电力负荷预测方法、装置及终端设备
CN120234775A (zh) 电力系统负荷的预测方法、装置、存储介质及电子设备
Keyan et al. Anomaly detection method of distribution network line loss based on hybrid clustering and LSTM
Zang et al. Short‐Term Wind Power Prediction Based on MVMD‐AVOA‐CNN‐LSTM‐AM
CN120955809A (zh) 一种融合电压稳定裕度与神经网络优化的配电网光伏可开放容量动态评估方法
CN110490220A (zh) 一种母线负荷辨识方法和系统
Randall et al. IoT based load forecasting for reliable integration of renewable energy sources
CN120109767A (zh) 一种基于配电网的分层分区信息监测方法及系统
CN120218980A (zh) 一种多模态的电力价格分析方法、系统、设备及存储介质
CN118884242A (zh) 一种电池能量状态预测方法、装置、设备及介质
Kushwah et al. Trend triplet based data clustering for eliminating nonlinear trend components of wind time series to improve the performance of statistical forecasting models
Guo et al. A Review of Wind Power Cluster Forecasting Techniques
Dong et al. Image-based with peak load ensemble prediction system for demand response in smart grid

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22933127

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22933127

Country of ref document: EP

Kind code of ref document: A1