Non-intrusive load decomposition method and system based on greedy algorithm
By using greedy algorithms and sample library fitting technology in non-invasive load detection, the problem of insufficient complexity and accuracy of load detection in the existing technology is solved, and the accurate judgment of industrial load and the adaptation to dynamic changes is achieved, thereby improving the improvement of the improvement of the problem of accurate judgment of industrial loads. The efficiency and accuracy of load decomposition are achieved.
Patent Information
- Application Number
- CN202411033166.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-22
AI Technical Summary
The existing non-invasive load detection and calculation methods are complex and cannot accurately judge the industrial load of the target user, and it is impossible to achieve accurate load electrical load decomposition when the industrial load type changes.
A non-invasive load decomposition method based on greedy algorithm is adopted. By establishing a sample library of industrial load equipment, the branch load curve of industrial users is obtained, and load decomposition is used to use greedy algorithm fitting to identify the operating status of the equipment to achieve accurate judgment.
It realizes accurate judgment of industrial load types, adapts to the dynamic changes of load curves, improves the accuracy and efficiency of load decomposition, reduces misjudgment, and ensures the reliability and accuracy of load decomposition results.
Smart Images

Figure CN119004129A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of power load monitoring, and in particular relates to a non-intrusive load decomposition method and system based on a greedy algorithm. Background Art
[0002] Existing load monitoring methods are mainly divided into two types: invasive and non-invasive. The invasive load monitoring method requires that each electrical device be equipped with an instrument with communication function, which will increase the cost of deploying and maintaining measuring instruments; the non-invasive load monitoring method requires that only one measuring instrument be installed at the user entrance of the power grid, and the collected power consumption information is analyzed through an algorithm to monitor the power consumption status of each electrical device under it. For large-scale deployment, the non-invasive load monitoring system can significantly reduce installation complexity and reduce maintenance costs.
[0003] CN202311119089.9A provides a non-invasive power load decomposition method and system, which relates to the field of power monitoring technology. The method includes analyzing the power difference between the total load power at the current moment and the total load power at the previous sampling time point according to the total load power of the household bus at the current moment; based on the power difference and the mutation power threshold, when it is determined that the load device on the household bus has changed, a real-time power change curve is generated starting from the load device change time point; the similarity between the real-time power change curve and the standard power change curve corresponding to each load device in the load device set is analyzed; when the similarity is greater than the similarity threshold, the load device corresponding to the similarity is determined to be a changing load, and the total load power is decomposed. This method has a large error in the recognition effect when the equipment adjusts the gear.
[0004] CN202311625381.8A A non-invasive load decomposition method and system, the method includes the following steps: obtaining load data of electrical equipment, establishing a load database based on the load data; collecting real-time power consumption data on the user side, and establishing a power load curve based on the real-time power consumption data, extracting features of the power load curve along the time direction, screening out continuous time segments where the extracted features are greater than a set value, and using them as loads to be identified; screening out the load data with the same features as the load to be identified from the load database, and using the electrical equipment and operating status corresponding to the load data as the identification results of the load to be identified. This method requires that the feature is greater than the set value before starting the identification algorithm, but the set value depends on experience, and varies greatly among different users, so there is a problem of missing users.
[0005] The existing non-invasive load detection calculation method is relatively complex and cannot make accurate and effective judgments based on the actual power consumption of the target user. Therefore, there is an urgent need for a non-invasive load decomposition algorithm that can accurately judge the type of industrial load used by the target user and accurately judge the power consumption of the target user's corresponding industrial load when the target user replaces the new industrial load. Summary of the invention
[0006] In order to solve the deficiencies in the prior art, the present invention provides a non-invasive load decomposition method based on a greedy algorithm. By establishing a sample library of loads of different industrial load equipment, the industrial user branch load curve is used as a detection curve, and the load decomposition is performed through greedy algorithm fitting.
[0007] The present invention implements a non-intrusive load decomposition method based on a greedy algorithm, the method comprising the following steps:
[0008] Step 1: Obtain equipment load samples of industrial users and establish an equipment load sample library;
[0009] Step 2: Collect branch load curves of industrial users;
[0010] Step 3: Screen the samples in the equipment load sample library as matching samples to form a matching sample library;
[0011] Step 4: Determine the operating state of the equipment branch load at each moment, and use a greedy algorithm to fit the load for different operating states of the equipment branch load. The operating state of the equipment branch load includes an increasing state, a decreasing state, and a stable state.
[0012] Step 5: Output load decomposition results.
[0013] The present invention further includes the following preferred embodiments:
[0014] In step 1, the equipment load sample includes the load curve within 1 minute from the start time, the start time of stable operation and the stop time of the industrial user equipment.
[0015] In step 2, the collection time of the industrial user branch load curve is greater than 10 minutes.
[0016] In step 3, the samples in the equipment load sample library are screened as matching samples to form a matching sample library, including:
[0017] Select the maximum load corresponding to the branch load curve obtained in step 2 as the screening criterion;
[0018] Traverse the equipment load sample library and select the load curves whose maximum load is lower than the screening standard as matching samples.
[0019] In step 4, the greedy algorithm is used to fit the load decomposition according to different operating states of the branch load, including:
[0020] Step 401: Setting the time of the industrial user branch load curve;
[0021] Step 402: Determine the operating status of the branch load at each moment;
[0022] Step 403: The branch load curve is fitted with the matching sample library at the corresponding time using a greedy algorithm and then load decomposition is performed.
[0023] In step 401, setting the time of the branch load curve of the industrial user includes setting the time of the branch load curve at intervals of 1 minute.
[0024] In step 402, determining the operating state of the branch load at each moment includes:
[0025] If the load increases by more than 10kW or the climbing rate is greater than 3kW per minute within 1 minute from the start of the status judgment moment, the branch load is judged to be in an increasing state; if the load decreases by more than 10kW or the climbing rate is less than -3kW per minute within 1 minute from the start of the status judgment moment, the branch load is judged to be in a decreasing state; otherwise, the branch load is judged to be in a stable state.
[0026] In step 403, the load decomposition is performed after the branch load curve is fitted with the matching sample library at the corresponding time by an algorithm, including: fitting the load curve at each operating state of the branch load with the equipment load curve at the corresponding time in the matching sample library by a greedy algorithm; calculating the average value of the branch load curve at different operating states of each device in the matching sample library, and determining the equipment load sample curve whose average value is less than and closest to the load value of the target detection curve as PS' nearest ; Calculate PS' at the matching time nearest The load value of the target detection curve is the difference between the load value. If the difference is less than 5%, the load decomposition result PS is output. result ={PS' nearest}; If the difference is greater than 5%, continue to select the equipment load sample curve that is smaller than and closest to the difference, and set the load decomposition result PS result =PS result ∪{PS' nearest}, and calculate PS at the matching time result The difference between the load value and the load value of the target detection curve is calculated until the difference is less than 5%, the cycle is stopped, and the load decomposition result is output. Otherwise, the equipment load sample curve that is less than and closest to the difference is continued to be superimposed.
[0027] At the same time, the present invention also provides a non-intrusive load decomposition system based on a greedy algorithm, including an equipment load sample library construction module, a branch load curve construction module, a matching sample library construction module, an algorithm fitting and load decomposition module;
[0028] The equipment load sample library construction module is used to obtain the load curves of each equipment at the startup time, stable operation time, and shutdown time;
[0029] The branch load curve building module is used to collect branch load curve data of target industrial users;
[0030] The matching sample library construction module is used to screen the load sample library and select all samples in the sample library that are lower than the maximum branch load as the matching sample library;
[0031] The algorithm fitting and load decomposition module is used to identify the moments when the branch load increases, decreases, and operates steadily, and to perform greedy algorithm fitting and load decomposition.
[0032] A terminal comprises a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of any one of the methods.
[0033] A computer-readable storage medium stores a computer program, which implements the steps of any one of the methods when executed by a processor.
[0034] The beneficial effect of the present invention is that, compared with the prior art, the present invention provides a non-intrusive load decomposition method based on a greedy algorithm. It has the following beneficial effects:
[0035] (1) The present invention establishes a sample library of loads of different industrial load equipment, takes the branch load curve of the industrial user as the detection curve, performs load decomposition through greedy algorithm fitting, and selects the industrial load combination that fits the branch load curve from the matching sample library, so as to achieve accurate judgment of the industrial load type used by the target user, and provide accurate and reliable calculation reference for the load decomposition of the industrial load used by the target user at each moment.
[0036] (2) The greedy algorithm used in the present invention has a fast calculation speed, is suitable for processing large amounts of data, and can complete load decomposition in real time or near real time. By building a superposition model, the data calculation of the detection curve and the sample load curve can be realized, which can not only conveniently screen out the output of the industrial load combination, but also provide an effective reference for the judgment of the type of industrial equipment used by the target user.
[0037] (3) The present invention can adapt to the dynamic changes of industrial user load curves, accurately capture the start and stop status of equipment and the stable operation stage through real-time monitoring and rapid response, and improve the accuracy and efficiency of load decomposition. By setting reasonable load curve screening standards and algorithm fitting thresholds (such as a difference of less than 5%), the reliability and accuracy of load decomposition results are ensured to avoid misjudgment caused by excessive differences between equipment load curves and actual branch loads. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 is a step diagram of a non-intrusive load decomposition method based on a greedy algorithm in an embodiment of the present invention;
[0039] Figure 2 is a branch load curve diagram of an industrial user collected in an embodiment of the present invention;
[0040] Figure 3 4 is a load decomposition result diagram in an embodiment of the present invention. DETAILED DESCRIPTION
[0041] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.
[0042] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.
[0043] The present invention provides a non-intrusive load decomposition method based on a greedy algorithm, such as Figure 1 FIG. 1 is a step diagram of a non-intrusive load decomposition method based on a greedy algorithm in an embodiment of the present invention, and the method includes the following steps:
[0044] Step 1: Obtain equipment load samples from industrial users and establish an equipment load sample library.
[0045] In the embodiment of the present invention, in step 1, the equipment load sample includes the load curve within 1 minute from the startup time, the start of stable operation time and the shutdown time of the industrial user equipment, which is recorded as {PS1,PS2,…,PS n}.
[0046] Step 2: Collect the branch load curve of industrial users within a set time period, recorded as PT(t).
[0047] In the embodiment of the present invention, a branch load curve of an industrial user in the 20-minute interval from 11:01 to 11:20 is collected, such as Figure 2 Shown is a branch load curve of an industrial user collected in this embodiment.
[0048] Step 3: Screen the samples in the equipment load sample library as matching samples to form a matching sample library.
[0049] In the embodiment of the present invention, in step 3, screening samples in the equipment load sample library as matching samples to form a matching sample library includes:
[0050] The maximum load corresponding to the branch load curve obtained in step 2, which is 3099kW, is selected as the screening criterion;
[0051] Traverse the equipment load sample library and select the load curves whose maximum load is lower than the screening standard as matching samples.
[0052] Step 4: Determine the operating state of the equipment branch load at each moment, and use a greedy algorithm to fit the load for different operating states of the equipment branch load. The operating state of the equipment branch load includes a rising state, a falling state, and a stable state.
[0053] In this embodiment, in step 4, the load decomposition is performed by using a greedy algorithm to fit the different operating states of the branch load, including:
[0054] Step 401: Setting the time of the branch load curve of the industrial user; in this embodiment, starting from 11:01, the time of the branch load curve is set at intervals of 1 minute.
[0055] Step 402: Determine the operating status of the branch load at each moment; if the load rises by more than 10kW or the ramp rate is greater than 3kW per minute within 1 minute from the moment of status judgment, that is, PT(t+1)>PT(t)+10 or dPT(t) / dt>3, then the branch load is determined to be in an increasing state; if the load drops by more than 10kW or the ramp rate is less than -3kW per minute within 1 minute from the moment of status judgment, that is, PT(t+1)<PT(t)-10 or dPT(t) / dt<-3, then the branch load is determined to be in a decreasing state; otherwise, the branch load is determined to be in a stable state.
[0056] In the embodiment of the present invention, taking 11:01 as an example, the load of the detection curve at this moment is 2389kW, and at 11:02 it is 2878kW. The load rises by more than 10kW within 1 minute. It is determined that 11:01 is in the rising state as an example for explanation:
[0057] The average value of the load curve at the startup time of each device in the matching sample library is calculated (one device corresponds to one average value). The average value calculated in the embodiment is shown in Table 1:
[0058] Table 1 Average value of load curve of each equipment at startup time
[0059] equipment Average value (kW) Unit 1# 2000 Unit 2# 150 Unit 3# 130 Unit 4# 1500 Unit 5# 650 Unit 6# 1000 Unit 7# 300 Unit 8# 100 Unit 9# 400 Unit 10# 60 Unit 11# 50 Unit 12# 80 Unit 13# 90 Unit 14# 420 Unit 15# 170
[0060] (2) Select the equipment curve whose average value is less than and closest to the load value of the test curve. The one that meets the requirements is unit 1#, 2000kW;
[0061] (3) Calculate the difference between the selected equipment load curve and the target detection curve, which is 389kW;
[0062] (4) The ratio of the calculated difference to the target detection curve is 16.28%, exceeding 5%;
[0063] (5) Continue to select the equipment curve whose average value is smaller than and closest to the difference. Assume that the one that meets the requirements is unit 2#, 150kW;
[0064] (6) The calculated load value of the superimposed equipment curve is 2150 kW, and the difference between the superimposed equipment load curve and the target detection curve is 239 kW;
[0065] (7) Calculate the ratio of the difference to the target detection curve, which is 10% and exceeds 5%;
[0066] (8) Continue to select the equipment curve whose average value is smaller than and closest to the difference. Assume that the one that meets the requirements is unit 3#, 130kW;
[0067] (9) The calculated load value of the superimposed equipment curve is 2280 kW, and the difference between the superimposed equipment load curve and the target detection curve is 109 kW;
[0068] (10) Calculate the ratio of the difference to the target detection curve, which is 4.5%, less than 5%;
[0069] (11) Output the load decomposition result at that moment. The decomposition result is shown in Table 2:
[0070] Table 2 Load decomposition results at startup time
[0071] Disassembled industrial equipment Power size at 11:01 (kW) Unit 1# 2000 Unit 2# 150 Unit 3# 130
[0072] Step 403: The branch load curve is fitted with the matching sample library at the corresponding time using a greedy algorithm and then load decomposition is performed.
[0073] In the embodiment of the present invention, in step S403, performing algorithm fitting on the branch load curve and the matching sample library at the corresponding time to perform load decomposition includes:
[0074] When the branch load increases, the load curve is fitted with the equipment load curve at the startup time in the matching sample library using a greedy algorithm; first, the average load value of the branch load curve at the startup time of each device in the matching sample library is calculated {Avg(PS'1), Avg(PS'2),…, Avg(PS' m )}, determine the equipment load sample curve whose average value is less than and closest to the load value of the detection curve as PS' nearest ; Calculate the difference between the load value of the determined equipment load curve and the load value of the target detection curve, and determine whether the difference is less than 5%, that is, PS' nearest (t)-PT(t)|<PT(t)*5%. If it is less than 5%, output load decomposition result PS result ={PS' nearest}; If it is greater than 5%, continue to select the device that is smaller than and closest to the difference, and set PS result ∪{PS' nearest} Set as the decomposition result; superimpose on the selected equipment curve until the difference between the superimposed load curve and the target detection curve is less than 5%, stop the cycle, and output the load decomposition result PS at that moment result , including the device type and the power at the corresponding moment.
[0075] When the branch load drops, the load curve is fitted with the equipment load curve at the shutdown time in the matching sample library using a greedy algorithm; first, the average load value of the branch load curve at the shutdown time of each device in the matching sample library is calculated {Avg(PS1), Avg(PS2),…, Avg(PS n )}, determine the equipment load sample curve whose average value is less than and closest to the load value of the detection curve as PS' nearest ; Calculate the difference between the load value of the determined equipment load curve and the load value of the target detection curve, and determine whether the difference is less than 5%, that is, PS' nearest (t)-PT(t)|<PT(t)*5%; if less than 5%, output load decomposition result PS result ={PS' nearest}; If it is greater than 5%, continue to select the device that is smaller than and closest to the difference, and set PS result ∪{PS' nearest} Set as the decomposition result; superimpose on the selected equipment curve until the difference between the superimposed load curve and the target detection curve is less than 5%, stop the cycle, and output the load decomposition result PS at that moment result , including the device type and the power at the corresponding moment.
[0076] At the moment of stable operation of branch load, the load curve is fitted with the equipment load curve at the stable operation moment in the matching sample library using a greedy algorithm; firstly, the average load value of the branch load curve at the shutdown moment of each equipment in the matching sample library is calculated {Avg(PS'1), Avg(PS'2),…, Avg(PS' m )}, determine the equipment load sample curve whose average value is less than and closest to the load value of the detection curve as PS' nearest ; Calculate the difference between the selected equipment load curve and the target detection curve, and determine whether the difference is less than 5%, that is, PS' nearest (t)-PT(t)|<PT(t)*5%; if less than 5%, output load decomposition result PS result ={PS' nearest}; If it is greater than 5%, continue to select the device that is smaller than and closest to the difference, and set PS result ∪{PS' nearest} Set as the decomposition result; superimpose on the selected equipment curve until the difference between the superimposed load curve and the target detection curve is less than 5%, stop the cycle, and output the load decomposition result PS at that moment result , including the device type and the power at the corresponding moment.
[0077] Step 5: Output the load decomposition results.
[0078] In the embodiment of the present invention, the time from 11:02 to 11:20 is continuously selected and processed according to the above algorithm, and the load decomposition results of all the time are output. The decomposition results are shown in Table 3:
[0079] Table 3 Load decomposition results at all times
[0080] time Disassembled industrial equipment Corresponding power size (kW) 11:01 Unit 1#, Unit 2#, Unit 3# 2000、150、130 11:02 Unit 1#, Unit 5#, Unit 3# 2000、650、130 11:03 Unit 1#, Unit 9#, Unit 10# 2000、400、60 11:04 Unit 1#, Unit 3#, Unit 11# 2000、130、50 11:05 Unit 1#, Unit 6# 2000、1000 11:06 Unit 1#, Unit 12# 2000、80 11:07 Unit 1#, Unit 7#, Unit 13# 2000、300、90 11:08 Unit 4#, Unit 13# 1500、90 11:09 Unit 4#, Unit 14# 1500、420 11:10 Unit 4#, Unit 8#, Unit 12# 1500、100、80 11:11 Unit 4#, Unit 7#, Unit 3# 1500、300、130 11:12 Unit 1#, Unit 7#, Unit 10# 2000、300、60 11:13 Unit 4#, Unit 9#, Unit 11# 1500、400、50 11:14 Unit 4#, Unit 2#, Unit 3# 1500、150、130 11:15 Unit 4#, Unit 9#, Unit 11# 1500、400、50 11:16 Unit 1#, Unit 5#, Unit 2#, Unit 12# 2000、650、150、80 11:17 Unit 4#, Unit 15#, Unit 11# 1500、170、50 11:18 Unit 4#, Unit 2#, Unit 3# 2000、150、130 11:19 Unit 1#, Unit 9#, Unit 2# 2000、400、150 11:20 Unit 4#, Unit 9# 1500、400
[0081] like Figure 3 The load decomposition result diagram in this embodiment is shown.
[0082] At the same time, the present invention also provides a non-intrusive load decomposition system based on a greedy algorithm, including an equipment load sample library construction module, a branch load curve construction module, a matching sample library construction module, an algorithm fitting and load decomposition module, characterized in that:
[0083] The equipment load sample library construction module is used to obtain the load curves of each equipment at the startup time, stable operation time, and shutdown time;
[0084] The branch load curve building module is used to collect branch load curve data of target industrial users;
[0085] The matching sample library construction module is used to screen the load sample library and select all samples in the sample library that are lower than the maximum branch load as the matching sample library;
[0086] The algorithm fitting and load decomposition module is used to identify the moments when the branch load increases, decreases, and operates steadily, and to perform greedy algorithm fitting and load decomposition.
[0087] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.
[0088] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.
[0089] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.
[0090] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.
[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A non-intrusive load decomposition method based on a greedy algorithm, characterized in that: The method comprises the following steps: Step 1: Obtain equipment load samples of industrial users and establish an equipment load sample library; Step 2: Collect branch load curves of industrial users; Step 3: Screen the samples in the equipment load sample library as matching samples to form a matching sample library; Step 4: Determine the operating state of the equipment branch load at each moment, and use a greedy algorithm to fit the load for different operating states of the equipment branch load. The operating state of the equipment branch load includes an increasing state, a decreasing state, and a stable state. Step 5: Output load decomposition results.
2. According to claim 1, a non-intrusive load decomposition method based on a greedy algorithm is characterized in that: In step 1, the equipment load sample includes the load curve within 1 minute from the start time, the start time of stable operation and the stop time of the industrial user equipment.
3. The non-intrusive load decomposition method based on a greedy algorithm according to claim 1, characterized in that: In step 2, the collection time of the industrial user branch load curve is greater than 10 minutes.
4. The non-intrusive load decomposition method based on a greedy algorithm according to claim 1, characterized in that: In step 3, the samples in the equipment load sample library are screened as matching samples to form a matching sample library, including: Select the maximum load corresponding to the branch load curve obtained in step 2 as the screening criterion; Traverse the equipment load sample library and select the load curves whose maximum load is lower than the screening standard as matching samples.
5. The non-intrusive load decomposition method based on a greedy algorithm according to claim 1 or 3, characterized in that: In step 4, the greedy algorithm is used to fit the load decomposition according to different operating states of the branch load, including: Step 401: Setting the time of the industrial user branch load curve; Step 402: Determine the operating status of the branch load at each moment; Step 403: The branch load curve is fitted with the matching sample library at the corresponding time using a greedy algorithm and then load decomposition is performed.
6. The non-intrusive load decomposition method based on a greedy algorithm according to claim 5, characterized in that: In step 401, setting the time of the industrial user branch load curve includes: Set the branch load curve time in 1 minute intervals.
7. The non-intrusive load decomposition method based on a greedy algorithm according to claim 6, characterized in that: In step 402, determining the operating state of the branch load at each moment includes: If the load rises by more than 10kW or the ramp rate is greater than 3kW per minute within 1 minute from the state judgment moment, the branch load is judged to be in the rising state; If the load drops by more than 10kW or the ramp rate is less than -3kW per minute within 1 minute from the state judgment moment, the branch load is judged to be in a decreasing state; Otherwise, it is determined that the branch load is in a stable state.
8. The non-intrusive load decomposition method based on a greedy algorithm according to claim 7, characterized in that: In step 403, performing algorithm fitting on the branch load curve and the matching sample library at the corresponding time to perform load decomposition includes: The load curves of the branch loads at each operating state are fitted with the equipment load curves at the corresponding time in the matching sample library using a greedy algorithm; Calculate the average load value of the branch load curve of each device in the matching sample library at different operating states, and determine the device load sample curve with an average value smaller than and closest to the target detection curve load value as PS' nearest ; Calculate the matching time PS' nearest The load value of the target detection curve is the difference between the load value of the target detection curve. If the difference is less than 5%, the load decomposition result PS is output. result ={PS' nearest }; If the difference is greater than 5%, continue to select the equipment load sample curve that is smaller than and closest to the difference, and set the load decomposition result PS result =PS result ∪{PS' nearest }, and calculate PS at the matching time result The difference between the load value and the load value of the target detection curve is calculated until the difference is less than 5%, the cycle is stopped, and the load decomposition result is output. Otherwise, the equipment load sample curve that is less than and closest to the difference is continued to be superimposed.
9. A non-intrusive load decomposition system based on a greedy algorithm using the method according to any one of claims 1 to 8, comprising a device load sample library construction module, a branch load curve construction module, a matching sample library construction module, an algorithm fitting and load decomposition module, characterized in that: The equipment load sample library construction module is used to obtain the load curves of each equipment at the startup time, stable operation time, and shutdown time; The branch load curve building module is used to collect branch load curve data of target industrial users; The matching sample library construction module is used to screen the load sample library and select all samples in the sample library that are lower than the maximum value of the branch load as the matching sample library; The algorithm fitting and load decomposition module is used to identify the moments when the branch load increases, decreases, and operates steadily, and to perform greedy algorithm fitting and load decomposition.
10. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-8.
11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.
Citation Information
Patent Citations
Non-intrusive load identification method based on regularization greedy forest algorithm
CN113408203A
Industrial electricity non-intrusive load decomposition method based on gradient lifting algorithm
CN114662563A
Photovoltaic generating capacity prediction method and device based on variational mode decomposition and ensemble learning
CN116629431A
Non-intrusive load monitoring method and terminal based on compressed sensing blind analysis
CN116953379A
Non-intrusive load decomposition method and system
CN117439077A
