CN117134470B - Sleep control method and related device of BMS battery management system - Google Patents
Sleep control method and related device of BMS battery management system Download PDFInfo
- Publication number
- CN117134470B CN117134470B CN202311399178.3A CN202311399178A CN117134470B CN 117134470 B CN117134470 B CN 117134470B CN 202311399178 A CN202311399178 A CN 202311399178A CN 117134470 B CN117134470 B CN 117134470B
- Authority
- CN
- China
- Prior art keywords
- power consumption
- data
- load
- target
- feature
- 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.)
- Active
Links
- 230000007958 sleep Effects 0.000 title claims abstract description 119
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000004458 analytical method Methods 0.000 claims abstract description 103
- 230000000737 periodic effect Effects 0.000 claims abstract description 103
- 230000005611 electricity Effects 0.000 claims abstract description 89
- 239000013598 vector Substances 0.000 claims abstract description 84
- 238000011217 control strategy Methods 0.000 claims abstract description 63
- 230000004927 fusion Effects 0.000 claims abstract description 52
- 230000005059 dormancy Effects 0.000 claims abstract description 50
- 238000006243 chemical reaction Methods 0.000 claims abstract description 29
- 238000000605 extraction Methods 0.000 claims abstract description 17
- 238000004364 calculation method Methods 0.000 claims abstract description 10
- 238000001514 detection method Methods 0.000 claims description 35
- 230000007246 mechanism Effects 0.000 claims description 27
- 238000005457 optimization Methods 0.000 claims description 24
- 238000007599 discharging Methods 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 16
- 238000013145 classification model Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 14
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000010606 normalization Methods 0.000 claims description 11
- 230000002068 genetic effect Effects 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 9
- 230000006978 adaptation Effects 0.000 claims description 8
- 230000035772 mutation Effects 0.000 claims description 6
- 230000002457 bidirectional effect Effects 0.000 claims description 5
- 230000008569 process Effects 0.000 description 13
- 230000006399 behavior Effects 0.000 description 11
- 230000008859 change Effects 0.000 description 7
- 238000005265 energy consumption Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
- H02J7/0063—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries with circuits adapted for supplying loads from the battery
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/425—Structural combination with electronic components, e.g. electronic circuits integrated to the outside of the casing
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
- H02J7/0068—Battery or charger load switching, e.g. concurrent charging and load supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
- H02J7/007—Regulation of charging or discharging current or voltage
- H02J7/00712—Regulation of charging or discharging current or voltage the cycle being controlled or terminated in response to electric parameters
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/425—Structural combination with electronic components, e.g. electronic circuits integrated to the outside of the casing
- H01M2010/4271—Battery management systems including electronic circuits, e.g. control of current or voltage to keep battery in healthy state, cell balancing
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Power Engineering (AREA)
- Manufacturing & Machinery (AREA)
- General Chemical & Material Sciences (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Chemical & Material Sciences (AREA)
- Electrochemistry (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to the technical field of battery management, and discloses a sleep control method and a related device of a BMS battery management system, which are used for realizing the intelligence of sleep control of the BMS battery management system and improving the accuracy of the sleep control. The method comprises the following steps: performing battery load calculation to obtain historical electricity load data; carrying out charge-discharge cycle extraction to obtain a plurality of power consumption cycle load data, and carrying out curve fitting to generate a plurality of power consumption cycle load curves; acquiring historical system power consumption data of the BMS battery management system, performing curve conversion, and generating a plurality of system periodic power consumption curves; extracting periodic characteristics to obtain a plurality of power load characteristic sets and a plurality of system power consumption characteristic sets; matching to obtain a target fusion feature set, and performing vector conversion to generate a plurality of target periodic feature vectors; and carrying out system dormancy control analysis through a system dormancy control analysis model to obtain a target system dormancy control strategy.
Description
Technical Field
The present invention relates to the field of battery management technologies, and in particular, to a sleep control method and a related device for a BMS battery management system.
Background
In the current society, the development of battery technology is one of the key factors for promoting the development of renewable energy, electric traffic, portable electronic devices and other fields. The design and optimization of a Battery Management System (BMS) are critical to improving battery performance, extending life, and ensuring safety. Sleep control is a key technology in a BMS to effectively manage energy consumption of a battery, improve energy efficiency of a system, and extend the life of the battery.
The traditional sleep control method only considers static power consumption analysis, but fails to fully consider dynamic change of power utilization period and time-varying change of system power consumption. The traditional method often ignores the behavior difference of the battery pack under different power utilization periods, and further leads to the reduction of the accuracy of the system sleep control strategy under certain periods.
Disclosure of Invention
The invention provides a sleep control method and a related device of a BMS battery management system, which are used for realizing the intelligence of sleep control of the BMS battery management system and improving the accuracy of the sleep control.
The first aspect of the present invention provides a sleep control method of a BMS battery management system, the sleep control method of the BMS battery management system comprising:
Acquiring historical electricity consumption data of a target battery pack through a preset BMS battery management system, and performing battery load calculation on the historical electricity consumption data to obtain historical electricity consumption load data;
performing charge-discharge cycle extraction on the historical electricity consumption load data to obtain a plurality of electricity consumption cycle load data, and performing curve fitting on the plurality of electricity consumption cycle load data to generate a plurality of electricity consumption cycle load curves;
acquiring historical system power consumption data of the BMS battery management system, extracting charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and performing curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves;
respectively extracting periodic characteristics of the plurality of power consumption periodic load curves to obtain a plurality of power consumption load characteristic sets, and extracting periodic characteristics of the plurality of system periodic power consumption curves to obtain a plurality of system power consumption characteristic sets;
matching the plurality of power consumption load feature sets with the plurality of system power consumption feature sets to obtain a plurality of target fusion feature sets in the same period, and performing vector conversion on the target fusion feature sets to generate a plurality of target period feature vectors;
And inputting the target period feature vectors into a preset system dormancy control analysis model to perform system dormancy control analysis, so as to obtain a target system dormancy control strategy.
With reference to the first aspect, in a first implementation manner of the first aspect of the present invention, the obtaining, by a preset BMS battery management system, historical electricity consumption data of a target battery pack, and performing battery load calculation on the historical electricity consumption data, to obtain historical electricity consumption load data includes:
acquiring historical electricity utilization data of a target battery pack through a preset BMS battery management system;
acquiring a voltage attribute tag and a current attribute tag, performing cluster center analysis on the voltage attribute tag through a preset data classification model to determine a corresponding voltage cluster center, and performing cluster center analysis on the current attribute tag through the data classification model to determine a corresponding current cluster center;
inputting the historical electricity consumption data into the data classification model, respectively calculating the data point distances between the historical electricity consumption data and the voltage clustering center and the current clustering center, and generating a corresponding target clustering result according to the data point distances;
Generating corresponding historical voltage data and historical current data according to the target clustering result, and performing time stamp alignment on the historical voltage data and the historical current data;
and calculating the product of the historical voltage data and the historical current data to obtain historical electricity load data.
With reference to the first aspect, in a second implementation manner of the first aspect of the present invention, the extracting the charge-discharge period of the historical power consumption load data to obtain a plurality of power consumption period load data, and performing curve fitting on the plurality of power consumption period load data to generate a plurality of power consumption period load curves, where the generating includes:
setting a first charging threshold and a first discharging threshold of the target battery pack, and performing charge-discharge state switching detection on the historical electricity load data according to the first charging threshold and the first discharging threshold to obtain a first charge-discharge state switching detection result;
based on the first charge-discharge state switching detection result, carrying out charge-discharge period data set division on the historical electricity load data to obtain a plurality of electricity period load data;
and acquiring first time stamp data corresponding to each power utilization period load data, and respectively performing curve fitting on the power utilization period load data by adopting a first spline interpolation function based on the first time stamp data to generate a plurality of power utilization period load curves.
With reference to the first aspect, in a third implementation manner of the first aspect of the present invention, the obtaining historical system power consumption data of the BMS battery management system, performing charge-discharge cycle extraction on the historical system power consumption data to obtain a plurality of system cycle power consumption data, performing curve conversion on the plurality of system cycle power consumption data, and generating a plurality of system cycle power consumption curves includes:
acquiring historical system power consumption data of the BMS battery management system, and setting a second charging threshold and a second discharging threshold of the BMS battery management system;
performing charge-discharge state switching detection on the historical system power consumption data according to the second charge threshold and the second discharge threshold to obtain a second charge-discharge state switching detection result;
according to the second charge-discharge state switching detection result, carrying out charge-discharge period data set division on the historical system power consumption data to obtain a plurality of system period power consumption data;
and acquiring second time stamp data corresponding to the power consumption data of each system period, and respectively performing curve fitting on the power consumption data of the plurality of system periods by adopting a second spline interpolation function based on the second time stamp data to generate a plurality of power consumption curves of the system period.
With reference to the first aspect, in a fourth implementation manner of the first aspect of the present invention, the extracting the cycle characteristics of the multiple power consumption cycle load curves to obtain multiple power consumption load characteristic sets, and extracting the cycle characteristics of the multiple system cycle power consumption curves to obtain multiple system power consumption characteristic sets respectively include:
defining a first periodic characteristic of an electricity periodic load curve, wherein the first periodic characteristic comprises a peak value, a valley value, an average value and a waveform factor;
based on the first periodic characteristics, curve characteristic points are extracted from the plurality of power consumption periodic load curves respectively, so that a plurality of first load curve characteristic points of each power consumption periodic load curve are obtained;
feature screening is carried out on the first load curve feature points to obtain a plurality of second load curve feature points of each power utilization period load curve, a plurality of corresponding power utilization load feature sets are generated according to the second load curve feature points of each power utilization period load curve, and the power utilization load feature sets are in one-to-one correspondence with the power utilization period load curves;
defining a second periodic characteristic of a system periodic power consumption curve, the second periodic characteristic comprising: average value, peak-valley difference, and power consumption fluctuation;
Based on the second periodic characteristics, extracting curve characteristic points of the system periodic power consumption curves respectively to obtain a plurality of first power consumption curve characteristic points of each system periodic power consumption curve;
and performing feature screening on the plurality of first power consumption curve feature points to obtain a plurality of second power consumption curve feature points of each system cycle power consumption curve, and generating a plurality of corresponding system power consumption feature sets according to the plurality of second power consumption curve feature points of each system cycle power consumption curve, wherein the system power consumption feature sets are in one-to-one correspondence with the system cycle power consumption curves.
With reference to the first aspect, in a fifth implementation manner of the first aspect of the present invention, the matching the multiple power load feature sets and the multiple system power consumption feature sets to obtain multiple target fusion feature sets in the same period, and performing vector conversion on the target fusion feature sets to generate multiple target period feature vectors, where the generating includes:
matching the plurality of power consumption characteristic sets and the plurality of system power consumption characteristic sets based on the first timestamp data and the second timestamp data to obtain the power consumption characteristic sets and the system power consumption characteristic sets of the same period;
Performing set fusion on the power consumption load feature set and the system power consumption feature set in the same period to generate a plurality of target fusion feature sets in the same period;
and carrying out serialization processing on the target fusion feature set to obtain a target fusion feature sequence, and carrying out vector conversion on the target fusion feature sequence to generate a plurality of target periodic feature vectors.
With reference to the first aspect, in a sixth implementation manner of the first aspect of the present invention, inputting the plurality of target period feature vectors into a preset system sleep control analysis model to perform system sleep control analysis, to obtain a target system sleep control policy, includes:
inputting the target period feature vectors into a preset system dormancy control analysis model, wherein the system dormancy control analysis model comprises a first layer control analysis network and a second layer strategy optimization network, the first layer control analysis network comprises a plurality of target control analysis networks, and each target control analysis network comprises a bidirectional long and short time memory network, a full connection layer, an attention mechanism layer and an inverse normalization layer;
receiving the target period feature vectors through a plurality of target control analysis networks in the first layer control analysis network respectively, and performing time sequence period feature operation on the target period feature vectors through the two-way long and short time memory network and the full connection layer to obtain time sequence hidden feature vectors;
Performing attention mechanism processing on the time sequence hidden feature vectors through the attention mechanism layer to generate corresponding attention mechanism feature vectors, performing control parameter analysis on the attention mechanism feature vectors through the inverse normalization layer, and outputting a system dormancy control parameter combination of each target control analysis network;
inputting a system dormancy control parameter combination of each target control analysis network into the second-layer strategy optimization network, and generating an initial control strategy population according to the system dormancy control parameter combination through a genetic algorithm in the second-layer strategy optimization network, wherein the initial control strategy population comprises a plurality of initial system dormancy control strategies;
calculating adaptation data of each initial system sleep control strategy respectively, and carrying out propagation, intersection and mutation operations on the plurality of initial system sleep control strategies according to the adaptation data to generate a plurality of candidate system sleep control strategies;
and optimally selecting the plurality of candidate system sleep control strategies to generate a target system sleep control strategy corresponding to the BMS battery management system.
A second aspect of the present invention provides a sleep control apparatus of a BMS battery management system, the sleep control apparatus of the BMS battery management system including:
The acquisition module is used for acquiring historical electricity utilization data of the target battery pack through a preset BMS battery management system, and carrying out battery load calculation on the historical electricity utilization data to obtain historical electricity utilization load data;
the fitting module is used for extracting charge and discharge cycles of the historical power utilization load data to obtain a plurality of power utilization cycle load data, and curve fitting the power utilization cycle load data to generate a plurality of power utilization cycle load curves;
the conversion module is used for acquiring historical system power consumption data of the BMS battery management system, extracting charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and performing curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves;
the extraction module is used for extracting the periodic characteristics of the plurality of power consumption periodic load curves to obtain a plurality of power consumption load characteristic sets, and extracting the periodic characteristics of the plurality of system periodic power consumption curves to obtain a plurality of system power consumption characteristic sets;
the matching module is used for matching the plurality of power load feature sets with the plurality of system power consumption feature sets to obtain a plurality of target fusion feature sets with the same period, and carrying out vector conversion on the target fusion feature sets to generate a plurality of target period feature vectors;
And the analysis module is used for inputting the plurality of target period feature vectors into a preset system dormancy control analysis model to carry out system dormancy control analysis so as to obtain a target system dormancy control strategy.
A third aspect of the present invention provides a sleep control apparatus of a BMS battery management system, comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the sleep control device of the BMS battery management system to perform the sleep control method of the BMS battery management system described above.
A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored therein, which when run on a computer, cause the computer to perform the sleep control method of the BMS battery management system described above.
According to the technical scheme provided by the invention, the historical electricity utilization data of the target battery pack can be accurately obtained through the clustering center analysis of the voltage and current attribute labels and the application of the data classification model. By adopting the double-threshold charge-discharge state switching detection method, the charge and discharge periods can be accurately extracted according to the set threshold, and the power consumption period load data and the system period power consumption data are subjected to curve fitting by adopting the first spline interpolation function and the second spline interpolation function, so that the system can reconstruct the power consumption load curve and the system power consumption curve more accurately, and the accuracy and the continuity of the data are improved. The electricity consumption behavior and the system power consumption condition in the period can be comprehensively known by defining and extracting a plurality of period characteristics of the electricity consumption period load curve and the system period power consumption curve, including peak values, valley values, average values, waveform factors, average values, peak-valley differences, power consumption fluctuation and the like. Through the process of matching and fusing the feature sets, the system can intelligently correspond the power load features and the system power consumption features to form a target fusion feature set in the same period, so that the periodic power consumption and system power consumption features are better captured. By inputting the target period feature vector to a preset system sleep control analysis model, a system sleep control strategy can be efficiently generated. The strategy is based on the historical data and the depth analysis of the feature set, and can be better adapted to the actual working state of the battery pack. The multi-level control analysis network and the strategy optimization network are utilized, periodic characteristics can be fully mined, sleep control parameters are optimized by adopting methods such as genetic algorithm and the like, and a more intelligent and strong-adaptability system sleep control strategy is generated, so that the sleep control intelligence of the BMS battery management system is realized, and the sleep control accuracy is improved.
Drawings
Fig. 1 is a schematic view illustrating an embodiment of a sleep control method of a BMS battery management system according to an embodiment of the present invention;
FIG. 2 is a flow chart of periodic feature extraction in an embodiment of the invention;
FIG. 3 is a flow chart of vector conversion according to an embodiment of the present invention;
FIG. 4 is a flow chart of system sleep control analysis in an embodiment of the invention;
fig. 5 is a schematic view illustrating an embodiment of a sleep control device of a BMS battery management system according to an embodiment of the present invention;
fig. 6 is a schematic view of an embodiment of a sleep control device of a BMS battery management system according to an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a sleep control method and a related device of a BMS battery management system, which are used for realizing the intelligence of sleep control of the BMS battery management system and improving the accuracy of the sleep control. The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For easy understanding, referring to fig. 1, an embodiment of a sleep control method of a BMS battery management system according to an embodiment of the present invention includes:
s101, acquiring historical electricity consumption data of a target battery pack through a preset BMS battery management system, and performing battery load calculation on the historical electricity consumption data to obtain historical electricity consumption load data;
it is to be understood that the execution subject of the present invention may be a sleep control device of a BMS battery management system, and may also be a terminal or a server, which is not limited herein. The embodiment of the invention is described by taking a server as an execution main body as an example.
Specifically, the server acquires historical electricity consumption data of the target battery pack through a preset BMS battery management system. This is to collect information about past power usage behavior of the battery. For example, the data may include information on the voltage, current, temperature, etc. of the battery. These historical electricity usage data will be used for battery load calculations. The purpose of this step is to calculate historical electrical load data from the voltage and current data. The electrical load of a battery is the product of voltage and current, which reflects the energy consumption of the battery at different points in time. For example, assume that a server has acquired voltage and current data for an electric vehicle over the past year. These data are recorded in units of minutes. And the server uses a preset data classification model to perform cluster center analysis on the voltage attribute tags and the current attribute tags. This will help the server determine cluster centers for voltage and current data, i.e. representative values for different behavior patterns or states. For example, the cluster center of the voltage attribute tags may include normal charge, normal travel, quick charge, and the like. The cluster center of the current attribute tag may represent a high-speed travel, a low-speed travel, a stationary state, or the like. The server inputs the historical electricity usage data into a data classification model. For each data point, the server calculates its distance from the voltage cluster center and the current cluster center. This will generate a target cluster result, i.e. which voltage and current pattern or state each data point corresponds to. For example, if voltage data at a certain time is closest to the center of the "normal charge" voltage cluster and current data is closest to the center of the "rest state" current cluster, then the target cluster result at that time is "normal charge+rest state". Based on the target cluster results, the server generates historical voltage data and historical current data, which have been aligned to the same time stamp. This is to ensure that the voltage and current data have the same time resolution as the electrical load data. The server calculates the product of the historical voltage data and the historical current data, thereby obtaining historical electricity load data. The power consumption load data reflect the energy consumption of the electric automobile at different time points, and are very important for making a sleep control strategy.
S102, extracting charge and discharge cycles of historical power utilization load data to obtain a plurality of power utilization cycle load data, and performing curve fitting on the plurality of power utilization cycle load data to generate a plurality of power utilization cycle load curves;
specifically, the server sets a first charge threshold and a first discharge threshold of the target battery pack. These thresholds are important parameters used to determine whether the battery is in a charged or discharged state. For example, the charge threshold may be set such that the battery voltage reaches a certain value, and the discharge threshold may be set such that the voltage is lower than a certain value. For example, the charge threshold is set to 50 volts and the discharge threshold is set to 40 volts. And performing charge-discharge state switching detection on the historical power load data by using the set thresholds. The purpose of this step is to determine when the battery switches from a charged state to a discharged state or vice versa in the historical electricity usage data. For example, when the voltage in the historical electricity usage data exceeds 50 volts, the battery is determined to be in a charged state; when the voltage is lower than 40 volts, the battery is determined to be in a discharge state. By detecting the voltage change, the switching point of the charge-discharge state can be determined. Based on the charge-discharge state switching detection result, the historical electricity load data is divided into a plurality of electricity utilization periods. Each cycle of power usage corresponds to a complete cycle of the battery between the charged and discharged states. For example, the point at which the battery is detected to switch from a charged state to a discharged state in the historical electricity usage data is the beginning of an electricity usage cycle. Correspondingly, the point at which the battery switches from the discharge state to the charge state is the end of one power utilization period. The data between these points is divided into load data for one power cycle. For each power cycle load data set, first time stamp data is obtained. Then, curve fitting is performed on the plurality of electricity periodic load data respectively using the first spline interpolation function. For example, for a particular power cycle, the first time stamp data indicates the start time of the cycle. By interpolation, a smooth power cycle load curve can be generated that reflects the charge and discharge of the battery over the cycle and its change over time.
S103, acquiring historical system power consumption data of the BMS battery management system, extracting charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and performing curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves;
it should be noted that historical system power consumption data of the BMS battery management system is obtained. These data record the energy consumption of the system at different points in time. The system power consumption data may include information such as current, voltage, and power of the entire system. For example, data of system power consumption in the past year is acquired from the BMS battery management system and recorded in units of minutes. In order to perform the charge-discharge state switching detection, a second charge threshold value and a second discharge threshold value need to be set. These thresholds are used to determine whether the system is in a charged or discharged state. For example, the second charge threshold is set to 100 kw and the second discharge threshold is set to 50 kw. And using a second charge and discharge threshold value to perform charge and discharge state switching detection on the historical system power consumption data. The purpose of this step is to determine when the system switches from a charged state to a discharged state or vice versa in the history data. For example, when historical system power consumption data exceeds 100 kilowatt-hours, the system is determined to be in a charged state; when the power consumption is below 50 kwh, the system is determined to be in a discharge state. By detecting the change in power consumption, the switching point of the charge-discharge state can be determined. Based on the second charge-discharge state switching detection result, the historical system power consumption data is divided into a plurality of system periods. Each system cycle corresponds to a complete cycle of the system between the charge and discharge states. For example, the point at which the system switches from a charged state to a discharged state is detected in the historical system power consumption data, i.e., the beginning of a system cycle. Correspondingly, the point at which the system switches from the discharge state to the charge state is the end of one system cycle. The data between these points is divided into power consumption data of one system cycle. For each system cycle power consumption dataset, second timestamp data is obtained. And then, respectively performing curve fitting on the plurality of system period power consumption data by using a second spline interpolation function. For example, for a particular system cycle, the second time stamp data indicates the start time of the cycle. By interpolation, a smooth system cycle power consumption curve can be generated that reflects the power consumption of the system over that cycle and its variation over time.
S104, respectively extracting periodic characteristics of a plurality of power consumption periodic load curves to obtain a plurality of power consumption load characteristic sets, and extracting periodic characteristics of a plurality of system periodic power consumption curves to obtain a plurality of system power consumption characteristic sets;
specifically, the server defines a first periodic characteristic of the power cycle load curve, including a peak value, a valley value, an average value, and a form factor. These features can be used to describe the performance and behavior of the load curve per power cycle. Then, the server extracts curve characteristic points of a plurality of power consumption period load curves according to the first period characteristics. The server finds these feature points within each cycle to describe the shape and characteristics of the load curve in more detail. Further, the server screens the extracted first load curve characteristic points to obtain second load curve characteristic points of each power utilization period load curve. These second feature points are more representative and can describe the key characteristics of the load curve more accurately. And generating a corresponding power utilization load characteristic set by the server according to the second load curve characteristic points of the load curve of each power utilization period. Each electricity utilization period load curve corresponds to the characteristic set one by one, so that the server can more comprehensively know the electricity utilization periodic behavior of the battery pack. The server defines a second periodic characteristic of the periodic power consumption curve of the system, including an average value, a peak-to-valley difference, and a power consumption fluctuation. These features are used to describe the performance and variation of the power consumption curve per system cycle. Then, the server performs curve feature point extraction on the plurality of system cycle power consumption curves according to the second cycle features to describe the shape and performance of the system power consumption curves in more detail. And then, the server screens the extracted characteristic points of the first power consumption curve to obtain the characteristic points of the second power consumption curve of each system period power consumption curve. These second feature points reflect key characteristics of power consumption in the system cycle, such as peak-to-valley differences and power consumption fluctuations. And generating a corresponding system power consumption characteristic set by the server according to the second power consumption curve characteristic points of each system period power consumption curve. Each system periodic power consumption curve corresponds to the characteristic set of the system, so that the server is helped to better understand the periodic power consumption behavior of the system.
S105, matching the plurality of power load feature sets with the plurality of system power consumption feature sets to obtain a plurality of target fusion feature sets in the same period, and performing vector conversion on the target fusion feature sets to generate a plurality of target period feature vectors;
specifically, the first timestamp data and the second timestamp data are used for matching a plurality of power consumption characteristic sets and a system power consumption characteristic set. Sets of power load and power consumption characteristics collected over the same period of time are found, which sets represent data over the same period. For example, if the first time stamp data represents a start time of the electrical load data and the second time stamp data represents a start time of the system power consumption data, then a match would find the electrical load and power consumption feature set collected over the same period of time. And carrying out set fusion on the matched power consumption load characteristic set and the matched system power consumption characteristic set in the same period. They are combined into a target fusion feature set that contains all the important feature information in the same period. For example, if the electrical load feature set includes peaks, valleys, averages, and form factors, and the system power consumption feature set includes average power consumption, peak-valley differences, and power consumption fluctuations, then the target fusion feature set will contain a combination of these features, such as peaks, average power consumption, and power consumption fluctuations, etc. And carrying out serialization treatment on the target fusion feature set, and converting the target fusion feature set into a target fusion feature sequence. This serialized feature set can be more conveniently stored, transmitted and further analyzed. For example, the target fusion feature set contains a plurality of feature values such as peak values, average power consumption, power consumption fluctuations, and the like. The characteristic values are organized into a sequence in a certain order, for example: peak value, average power consumption, power consumption fluctuation ]. And carrying out vector conversion on the target fusion characteristic sequence to generate a plurality of target periodic characteristic vectors. These feature vectors are key tools to describe system performance and periodic behavior and can be used for further analysis and optimization of sleep control strategies. For example, the target fusion feature sequence [ peak, average power consumption, power consumption fluctuation ] may be converted into a feature vector [0.95, 500 kw, 0.2], where each value corresponds to a specific value of a feature.
S106, inputting the characteristic vectors of the multiple target periods into a preset system dormancy control analysis model to carry out system dormancy control analysis, and obtaining a target system dormancy control strategy.
Specifically, the generated target periodic feature vector will become the input of the system sleep control analysis model. These vectors will contain key features of power cycle and system power consumption for intelligent analysis and optimization. The first layer control analysis network is composed of a plurality of target control analysis networks, each network comprising the following components: a Bi-directional long-short-term memory network (Bi-LSTM) is used to process the input target period feature vector and capture the timing period feature. This helps to understand the dynamic changes in battery and system behavior; the fully connected layer is used to correlate the output of the Bi-LSTM with the control parameters. These parameters include charge threshold, discharge threshold, etc.; the attention mechanism layer may determine which points in time data are most important for the current control decision to generate an attention mechanism feature vector; the inverse normalization layer translates the output of the attention mechanism into a specific combination of control parameters that will be used to control the sleep behavior of the battery pack during the current cycle. Next, the second tier policy optimization network further optimizes the control parameters generated by the first tier to generate an optimal system sleep control policy. An initial control strategy population is generated, the initial control strategy population comprising a plurality of initial system sleep control strategies, each strategy comprising a set of control parameters. For each initial system sleep control strategy, its fitness data is calculated. The fitness data evaluates indexes in aspects of battery pack performance, battery life and the like according to historical electricity utilization data, periodic feature extraction, system power consumption data and the like. And according to the fitness data, using optimization methods such as genetic algorithm and the like to carry out propagation, crossover and mutation operations on the initial system dormancy control strategy, and generating a plurality of candidate system dormancy control strategies. And selecting a strategy with highest fitness from the candidate system sleep control strategies as a target system sleep control strategy of the BMS battery management system. For example, assume that a server collects historical power consumption data of a system and system power consumption data, and then performs data preprocessing and cycle characteristic extraction. The server inputs the target periodic feature vector into a system sleep control analysis model, which includes a first layer control analysis network and a second layer policy optimization network. The first layer control analysis network utilizes the Bi-LSTM, full connection layer, attention mechanism layer, and inverse normalization layer to extract timing cycle characteristics and generate control parameter combinations. And the second-layer strategy optimization network optimizes the control parameters by using a genetic algorithm to generate an optimal system dormancy control strategy. The server selects the sleep control strategy with the highest fitness that will be applied to the electric vehicle energy storage system for optimal energy efficiency and battery performance. By the method, the server can realize intelligent battery management, maximally utilize renewable energy sources and protect the service life of the battery. This helps to promote the sustainable development of renewable energy systems.
According to the embodiment of the invention, the historical electricity utilization data of the target battery pack can be accurately obtained through the clustering center analysis of the voltage and current attribute labels and the application of the data classification model. By adopting the double-threshold charge-discharge state switching detection method, the charge and discharge periods can be accurately extracted according to the set threshold, and the power consumption period load data and the system period power consumption data are subjected to curve fitting by adopting the first spline interpolation function and the second spline interpolation function, so that the system can reconstruct the power consumption load curve and the system power consumption curve more accurately, and the accuracy and the continuity of the data are improved. The electricity consumption behavior and the system power consumption condition in the period can be comprehensively known by defining and extracting a plurality of period characteristics of the electricity consumption period load curve and the system period power consumption curve, including peak values, valley values, average values, waveform factors, average values, peak-valley differences, power consumption fluctuation and the like. Through the process of matching and fusing the feature sets, the system can intelligently correspond the power load features and the system power consumption features to form a target fusion feature set in the same period, so that the periodic power consumption and system power consumption features are better captured. By inputting the target period feature vector to a preset system sleep control analysis model, a system sleep control strategy can be efficiently generated. The strategy is based on the historical data and the depth analysis of the feature set, and can be better adapted to the actual working state of the battery pack. The multi-level control analysis network and the strategy optimization network are utilized, periodic characteristics can be fully mined, sleep control parameters are optimized by adopting methods such as genetic algorithm and the like, and a more intelligent and strong-adaptability system sleep control strategy is generated, so that the sleep control intelligence of the BMS battery management system is realized, and the sleep control accuracy is improved.
In a specific embodiment, the process of executing step S101 may specifically include the following steps:
(1) Acquiring historical electricity utilization data of a target battery pack through a preset BMS battery management system;
(2) Acquiring a voltage attribute tag and a current attribute tag, performing cluster center analysis on the voltage attribute tag through a preset data classification model to determine a corresponding voltage cluster center, and performing cluster center analysis on the current attribute tag through the data classification model to determine a corresponding current cluster center;
(3) Inputting historical electricity utilization data into a data classification model, respectively calculating the data point distances of the historical electricity utilization data, a voltage clustering center and a current clustering center, and generating a corresponding target clustering result according to the data point distances;
(4) Generating corresponding historical voltage data and historical current data according to the target clustering result, and performing time stamp alignment on the historical voltage data and the historical current data;
(5) And calculating the product of the historical voltage data and the historical current data to obtain historical electricity load data.
Specifically, the server acquires historical electricity consumption data of the target battery pack through a preset BMS battery management system. These data include information such as the voltage, current, and time stamp of the battery. Then, attribute tags of voltage and current are defined. These tags may include voltage peaks, current averages, voltage valleys, etc. The selection of these attribute tags will facilitate subsequent data analysis. And performing cluster center analysis on the voltage attribute tags and the current attribute tags by using a preset data classification model. This model may be a clustering algorithm, such as K-means clustering or hierarchical clustering, for dividing the voltage and current data into different clusters and determining the cluster center for each cluster. And processing the historical electricity consumption data, and calculating the distance between each data point and the voltage clustering center and the current clustering center. This may be done using euclidean distance or other distance metric methods. Then, each data point is distributed to the nearest voltage and current clustering center, and a target clustering result is generated. Historical electricity usage data is divided into different voltage and current categories based on the target clustering results. This will generate a time series of historical voltage and current data, each point in time corresponding to a cluster center. Since historical electricity data, voltage data, and current data have different time stamps, time stamp alignment is required. This may be done by interpolation or other time series alignment methods to ensure that the data is synchronized at the same point in time. And multiplying the historical voltage data and the historical current data to obtain historical electricity load data. This process reflects the actual power usage of the battery pack, which will be used for subsequent sleep control analysis. For example, assume that the server collects historical voltage and current data for the battery pack, along with a corresponding timestamp. The server defines attribute tags for voltage and current, including voltage peaks, current averages, and current valleys. Then, the server performs cluster center analysis on the voltage and current data by using a K-means clustering algorithm, and a cluster center of each attribute is determined. The server then calculates the distance between each historical data point and the voltage and current cluster center and assigns each data point to a corresponding cluster center based on the minimum distance. Thus, the server generates a target cluster result, and determines the voltage and current category to which each historical data point belongs. The server then generates a time series of historical voltage and current data from the target cluster results and performs a time stamp alignment to ensure that they are synchronized at the same point in time. The server multiplies the historical voltage and current data to obtain historical electricity load data. The data reflects the electricity load condition of the battery pack of the electric automobile at different time points, and can be used for further sleep control analysis and strategy formulation. This helps to optimize the use of the battery and extend its life.
In a specific embodiment, the process of executing step S102 may specifically include the following steps:
(1) Setting a first charging threshold and a first discharging threshold of a target battery pack, and performing charge-discharge state switching detection on historical power utilization load data according to the first charging threshold and the first discharging threshold to obtain a first charge-discharge state switching detection result;
(2) Based on the first charge-discharge state switching detection result, carrying out charge-discharge period data set division on historical power utilization load data to obtain a plurality of power utilization period load data;
(3) And acquiring first timestamp data corresponding to each power utilization period load data, and respectively performing curve fitting on the power utilization period load data by adopting a first spline interpolation function based on the first timestamp data to generate a plurality of power utilization period load curves.
Specifically, the server determines a first charge threshold and a first discharge threshold for the target battery pack. These thresholds may be set according to the specifications and performance requirements of the battery. The first charge threshold is typically used to determine when the battery begins to charge, and the first discharge threshold is used to determine when to begin discharging. And carrying out state of charge switching detection on each data point by using the historical electricity load data. This may be accomplished by comparing the electrical load of each data point to a first charge threshold and a first discharge threshold. The battery is considered to be in a charged state when the electrical load exceeds a first charge threshold, and is considered to be in a discharged state when the electrical load is below a first discharge threshold. The detection result generates a binary sequence representing the charge and discharge states of the battery. Based on the charge-discharge state switching detection result, the historical electricity load data is divided into a plurality of charge cycles and discharge cycles. These periods represent different charge and discharge processes of the battery during different time periods. Typically, when a switch from a charged state to a discharged state (or vice versa) is detected, one charge-discharge cycle ends and the other begins. And for each charge-discharge period, acquiring corresponding first timestamp data. These timestamp data identify the starting point in time for each cycle that can be used for subsequent curve fitting. Based on the first timestamp data, curve fitting is performed on the electrical load data for each charge cycle and discharge cycle using a first spline interpolation function. The first spline interpolation is a common interpolation method that can generate a smooth curve to better represent the change in electrical load over time. For example, assume that the server sets the first charge threshold to 100kW and the first discharge threshold to 10kW. Then, the server performs charge-discharge state switching detection using historical electricity load data, such as load data recorded every minute. When the load exceeds 100kW, the server considers the battery to be in a charged state; when the load is below 10kW, the server considers the battery to be in a discharged state. By detecting, the server generates a binary sequence of charge and discharge states. Then, the server divides the historical electricity load data into a plurality of charging periods and discharging periods according to the charging and discharging state switching detection result. Each time a switch from a charged state to a discharged state (or vice versa) is detected, one cycle is ended and another cycle is started. For curve fitting, the server obtains first time stamp data for each charge cycle and discharge cycle, representing their starting time points. Then, the server performs curve fitting on the electricity load data of each period by using a first spline interpolation function to generate load curves of the charging period and the discharging period.
In a specific embodiment, the process of executing step S103 may specifically include the following steps:
(1) Acquiring historical system power consumption data of the BMS battery management system, and setting a second charging threshold and a second discharging threshold of the BMS battery management system;
(2) Performing charge-discharge state switching detection on the historical system power consumption data according to the second charge threshold and the second discharge threshold to obtain a second charge-discharge state switching detection result;
(3) According to the second charge-discharge state switching detection result, carrying out charge-discharge period data set division on the historical system power consumption data to obtain a plurality of system period power consumption data;
(4) And acquiring second time stamp data corresponding to each system period power consumption data, and respectively performing curve fitting on the plurality of system period power consumption data by adopting a second spline interpolation function based on the second time stamp data to generate a plurality of system period power consumption curves.
Specifically, the server acquires historical system power consumption data from the BMS battery management system. These data may include a record of the power consumption of the battery pack at various points in time, typically expressed in terms of power (or current and voltage) over time. A second charge threshold and a second discharge threshold of the BMS battery management system are determined. These thresholds are typically set based on requirements of system performance and security. The second charge threshold is used to determine when the battery is in a charged state and the second discharge threshold is used to determine when the battery is in a discharged state. And using historical system power consumption data to perform charge-discharge state switching detection on each data point. This may be achieved by comparing the system power consumption of each data point with a second charge threshold and a second discharge threshold. When the system power consumption exceeds the second charge threshold, the system is considered to be in a charged state, and when the system power consumption is below the second discharge threshold, the system is considered to be in a discharged state. The detection result will generate a binary sequence representing the charge and discharge state of the system. Based on the charge-discharge state switching detection result, the historical system power consumption data is divided into a plurality of charge cycles and discharge cycles. Each time a switch from a charged state to a discharged state (or vice versa) is detected, one cycle is ended and another cycle is started. And for each charging period and discharging period, acquiring corresponding second timestamp data. These timestamp data identify the starting point in time for each cycle that can be used for subsequent curve fitting. And performing curve fitting on the system power consumption data of each charging period and discharging period by using a second spline interpolation function based on the second timestamp data. The second spline interpolation is a common interpolation method that can generate a smooth curve to better represent the change in system power consumption over time. For example, assume that the server sets the second charge threshold to 50kW and the second discharge threshold to 10kW. Then, the server performs charge-discharge state switching detection using historical system power consumption data, for example, power consumption data recorded every hour. When the power consumption exceeds 50kW, the server considers the system to be in a charging state; when the power consumption is below 10kW, the server considers the system to be in a discharge state. By detecting, the server generates a binary sequence of charge and discharge states. Then, the server divides the historical system power consumption data into a plurality of charging periods and discharging periods according to the charging and discharging state switching detection result. Each time a switch from a charged state to a discharged state (or vice versa) is detected, one cycle is ended and another cycle is started. For curve fitting, the server obtains second time stamp data for each charge cycle and discharge cycle, representing their starting points in time. And then, the server performs curve fitting on the system power consumption data of each period by using a second spline interpolation function to generate power consumption curves of the charging period and the discharging period.
In a specific embodiment, as shown in fig. 2, the process of executing step S104 may specifically include the following steps:
s201, defining a first periodic characteristic of a power consumption periodic load curve, wherein the first periodic characteristic comprises a peak value, a valley value, an average value and a waveform factor;
s202, based on the first periodic characteristics, extracting curve characteristic points of a plurality of power consumption periodic load curves respectively to obtain a plurality of first load curve characteristic points of each power consumption periodic load curve;
s203, performing feature screening on the first load curve feature points to obtain a plurality of second load curve feature points of each power utilization period load curve, and generating a plurality of corresponding power utilization load feature sets according to the second load curve feature points of each power utilization period load curve, wherein the power utilization load feature sets correspond to the power utilization period load curves one by one;
s204, defining a second periodic characteristic of a periodic power consumption curve of the system, wherein the second periodic characteristic comprises: average value, peak-valley difference, and power consumption fluctuation;
s205, based on the second periodic characteristics, extracting curve characteristic points of a plurality of system periodic power consumption curves respectively to obtain a plurality of first power consumption curve characteristic points of each system periodic power consumption curve;
S206, feature screening is carried out on the first power consumption curve feature points to obtain second power consumption curve feature points of each system period power consumption curve, and corresponding system power consumption feature sets are generated according to the second power consumption curve feature points of each system period power consumption curve and correspond to the system period power consumption curves one by one.
In particular, the server defines a first periodic characteristic of the power cycle load curve. Typically including peaks, valleys, averages and form factors. These features are used to describe the basic characteristics of the power cycle load curve. And for a plurality of electricity consumption period load curves, extracting curve characteristic points of each electricity consumption period load curve based on the first period characteristic. For example, peak points, valley points, average points, etc. in the curve may be found. These points are used to represent key features of the curve. And carrying out feature screening on the extracted curve feature points. Screening includes removing noise or irrelevant feature points to ensure that the retained feature points accurately characterize the power cycle load curve. This step helps to improve the accuracy and usability of the features. And generating an electricity load characteristic set for each electricity period load curve based on the screened characteristic points. This feature set contains the important features of the periodic load curve, in one-to-one correspondence with the curve. Likewise, a second periodic characteristic of the system periodic power consumption curve is defined. These characteristics may include averages, peak-to-valley differences, power consumption fluctuations, etc. that characterize the periodic power consumption curve of the system. And for a plurality of system cycle power consumption curves, extracting curve characteristic points of each system cycle power consumption curve based on the second cycle characteristic. These points may represent important characteristics of the system power consumption curve. And (3) performing feature screening on the extracted curve feature points to remove unnecessary or irrelevant feature points so as to improve the accuracy of the features. And generating a system power consumption characteristic set for each system period power consumption curve based on the screened characteristic points. This feature set contains key features of the periodic power consumption curve, which are in one-to-one correspondence with the curve.
In a specific embodiment, as shown in fig. 3, the process of executing step S105 may specifically include the following steps:
s301, matching a plurality of power consumption characteristic sets and a plurality of system power consumption characteristic sets based on first timestamp data and second timestamp data to obtain the power consumption characteristic sets and the system power consumption characteristic sets of the same period;
s302, carrying out set fusion on the power load feature set and the system power consumption feature set in the same period to generate a plurality of target fusion feature sets in the same period;
s303, carrying out serialization processing on the target fusion feature set to obtain a target fusion feature sequence, and carrying out vector conversion on the target fusion feature sequence to generate a plurality of target periodic feature vectors.
In particular, the server ensures that the first timestamp data and the second timestamp data can correctly match the electrical load feature set and the system power consumption feature set. Time alignment and synchronization is performed during the data collection and processing stages to ensure that both types of feature sets have the same time stamp or can be mapped onto the same time period. The first time stamp data is associated with a power usage load feature set and the second time stamp data is associated with a system power consumption feature set. Ensuring that they are all associated with the same time period for subsequent set fusion and feature extraction. And fusing the power consumption characteristic set and the system power consumption characteristic set to generate a plurality of target fusion characteristic sets in the same period. The set fusion may employ different strategies depending on the particular application requirements. Including simply linking features together, weighted fusion or using other statistical methods. The target fusion feature set is serialized and converted into a serialized data structure for subsequent processing and storage. Serialization can take a variety of formats, such as JSON, XML, or binary formats, depending on the application requirements of the server and the manner in which the data is stored. And carrying out vector conversion on the serialized target fusion feature sequence to generate a plurality of target periodic feature vectors. Vector conversion involves mapping data in a sequence into a low-dimensional feature vector space for subsequent analysis and modeling. This may be accomplished using dimensionality reduction techniques (such as principal component analysis) or other vectorization methods. For example, assume that first timestamp data is matched to a power usage load feature set and second timestamp data is matched to a system power consumption feature set. And carrying out weighted fusion on the power consumption characteristic set and the system power consumption characteristic set to obtain a plurality of target fusion characteristic sets in the same period. Then, the target fusion feature set is serialized into JSON formatted data. Vector conversion of the serialized data generates a plurality of target periodic feature vectors that can be used for further analysis, modeling, or decision making.
In a specific embodiment, as shown in fig. 4, the process of executing step S106 may specifically include the following steps:
s401, inputting a plurality of target period feature vectors into a preset system dormancy control analysis model, wherein the system dormancy control analysis model comprises a first layer control analysis network and a second layer strategy optimization network, the first layer control analysis network comprises a plurality of target control analysis networks, and each target control analysis network comprises a bidirectional long and short time memory network, a full connection layer, an attention mechanism layer and an inverse normalization layer;
s402, respectively receiving a plurality of target periodic feature vectors through a plurality of target control analysis networks in a first layer control analysis network, and carrying out time sequence periodic feature operation on the target periodic feature vectors through a bidirectional long-short-term memory network and a full-connection layer to obtain time sequence hidden feature vectors;
s403, performing attention mechanism processing on the time-lapse hidden feature vectors through an attention mechanism layer to generate corresponding attention mechanism feature vectors, performing control parameter analysis on the attention mechanism feature vectors through an inverse normalization layer, and outputting a system dormancy control parameter combination of each target control analysis network;
S404, inputting the system dormancy control parameter combination of each target control analysis network into a second-layer strategy optimization network, and generating an initial control strategy population according to the system dormancy control parameter combination through a genetic algorithm in the second-layer strategy optimization network, wherein the initial control strategy population comprises a plurality of initial system dormancy control strategies;
s405, respectively calculating the adaptation data of each initial system sleep control strategy, and carrying out propagation, intersection and mutation operations on a plurality of initial system sleep control strategies according to the adaptation data to generate a plurality of candidate system sleep control strategies;
and S406, optimizing and selecting the plurality of candidate system sleep control strategies, and generating a target system sleep control strategy corresponding to the BMS battery management system.
Specifically, the server inputs a plurality of target periodic feature vectors into a system sleep control analysis model. These feature vectors are derived from the previous steps and contain important information about the power load and the power consumption of the system. Each target control analysis network will receive a target periodic feature vector. The first layer control analysis network is a key component and comprises a plurality of target control analysis networks, and each target control analysis network has a plurality of key layers. The target control analysis network firstly uses a Bi-directional long-short-time memory network (Bi-LSTM) to perform time sequence periodic characteristic operation on the input target periodic characteristic vector. This facilitates the timing relationship between the network capture features. The fully connected layer is used for further processing the features, extracting higher-level features related to system sleep control, and the attention mechanism layer is used for processing the features to generate attention mechanism feature vectors. This helps the network focus on important features to better control system dormancy, and the anti-normalization layer performs control parameter analysis on the attention mechanism feature vectors to generate a system dormancy control parameter combination for each target control analysis network. These parameters will be used to control the dormancy of the battery system. The second layer policy optimization network uses genetic algorithm to generate initial control policy population according to the system dormancy control parameter combination. The initial control strategy population includes a plurality of initial system sleep control strategies that analyze the output of the network based on the first layer control. For each initial system sleep control strategy, its fitness data needs to be calculated. These data reflect the performance of the policy in the actual system. And carrying out propagation, crossing and mutation operations on the plurality of initial system sleep control strategies according to the fitness data so as to generate a plurality of candidate system sleep control strategies. And selecting an optimal strategy from the plurality of candidate system sleep control strategies as a target system sleep control strategy of the BMS battery management system. This strategy will control the sleep state of the battery pack in real time according to the system requirements to improve energy efficiency, extend life and ensure safety. For example, assuming that the server inputs historical data and periodic feature vectors, a set of control parameters is generated through the computation of the first layer control analysis network and the second layer policy optimization network. These parameters tell the battery when to enter sleep mode and when to wake up to meet power demand. Through continuous iteration and optimization, the server finally obtains an efficient system dormancy control strategy, and the service life of the battery can be prolonged and the reliability can be ensured under the condition of not sacrificing the performance.
The above describes a sleep control method of a BMS battery management system according to an embodiment of the present invention, and the following describes a sleep control device of a BMS battery management system according to an embodiment of the present invention, referring to fig. 5, and one embodiment of the sleep control device of a BMS battery management system according to an embodiment of the present invention includes:
the acquiring module 501 is configured to acquire historical electricity consumption data of a target battery pack through a preset BMS battery management system, and perform battery load calculation on the historical electricity consumption data to obtain historical electricity consumption load data;
the fitting module 502 is configured to perform charge-discharge cycle extraction on the historical power consumption load data to obtain a plurality of power consumption cycle load data, and perform curve fitting on the plurality of power consumption cycle load data to generate a plurality of power consumption cycle load curves;
a conversion module 503, configured to obtain historical system power consumption data of the BMS battery management system, extract charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and perform curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves;
the extracting module 504 is configured to perform periodic feature extraction on the multiple power consumption periodic load curves to obtain multiple power consumption load feature sets, and perform periodic feature extraction on the multiple system periodic power consumption curves to obtain multiple system power consumption feature sets;
The matching module 505 is configured to match the multiple power load feature sets and the multiple system power consumption feature sets to obtain multiple target fusion feature sets in the same period, and perform vector conversion on the target fusion feature sets to generate multiple target period feature vectors;
and the analysis module 506 is configured to input the multiple target period feature vectors into a preset system sleep control analysis model to perform system sleep control analysis, so as to obtain a target system sleep control policy.
Through the cooperative cooperation of the components, the beneficial effects are that
The sleep control device of the BMS battery management system in the embodiment of the present invention is described in detail from the point of view of the modularized functional entity in fig. 5 above, and the sleep control apparatus of the BMS battery management system in the embodiment of the present invention is described in detail from the point of view of the hardware processing.
Fig. 6 is a schematic structural diagram of a sleep control device of a BMS battery management system according to an embodiment of the present invention, where the sleep control device 600 of the BMS battery management system may have a relatively large difference according to configuration or performance, and may include one or more processors (central processing units, CPU) 610 (e.g., one or more processors) and a memory 620, and one or more storage media 630 (e.g., one or more mass storage devices) storing application programs 633 or data 632. Wherein the memory 620 and the storage medium 630 may be transitory or persistent storage. The program stored in the storage medium 630 may include one or more modules (not shown), each of which may include a series of instruction operations in the sleep control device 600 of the BMS battery management system. Still further, the processor 610 may be configured to communicate with the storage medium 630 to perform a series of instruction operations in the storage medium 630 on the sleep control device 600 of the BMS battery management system.
The sleep control device 600 of the BMS battery management system may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input/output interfaces 660, and/or one or more operating systems 631, such as Windows Server, mac OS X, unix, linux, freeBSD, and the like. It will be appreciated by those skilled in the art that the sleep control device structure of the BMS battery management system shown in fig. 6 does not constitute a limitation of the sleep control device of the BMS battery management system, and may include more or less components than those illustrated, or may combine certain components, or may have different arrangements of components.
The present invention also provides a sleep control apparatus of a BMS battery management system, which includes a memory and a processor, wherein the memory stores computer readable instructions that, when executed by the processor, cause the processor to perform the steps of the sleep control method of the BMS battery management system in the above embodiments.
The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and which may also be a volatile computer-readable storage medium, having instructions stored therein that, when executed on a computer, cause the computer to perform the steps of the sleep control method of the BMS battery management system.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. A sleep control method of a BMS battery management system, the sleep control method of the BMS battery management system comprising:
acquiring historical electricity consumption data of a target battery pack through a preset BMS battery management system, and performing battery load calculation on the historical electricity consumption data to obtain historical electricity consumption load data;
performing charge-discharge cycle extraction on the historical electricity consumption load data to obtain a plurality of electricity consumption cycle load data, and performing curve fitting on the plurality of electricity consumption cycle load data to generate a plurality of electricity consumption cycle load curves;
acquiring historical system power consumption data of the BMS battery management system, extracting charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and performing curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves;
Respectively extracting periodic characteristics of the plurality of power consumption periodic load curves to obtain a plurality of power consumption load characteristic sets, and extracting periodic characteristics of the plurality of system periodic power consumption curves to obtain a plurality of system power consumption characteristic sets;
matching the plurality of power consumption load feature sets with the plurality of system power consumption feature sets to obtain a plurality of target fusion feature sets in the same period, and performing vector conversion on the target fusion feature sets to generate a plurality of target period feature vectors;
inputting the multiple target period feature vectors into a preset system dormancy control analysis model to carry out system dormancy control analysis, so as to obtain a target system dormancy control strategy; the method specifically comprises the following steps: inputting the target period feature vectors into a preset system dormancy control analysis model, wherein the system dormancy control analysis model comprises a first layer control analysis network and a second layer strategy optimization network, the first layer control analysis network comprises a plurality of target control analysis networks, and each target control analysis network comprises a bidirectional long and short time memory network, a full connection layer, an attention mechanism layer and an inverse normalization layer; receiving the target period feature vectors through a plurality of target control analysis networks in the first layer control analysis network respectively, and performing time sequence period feature operation on the target period feature vectors through the two-way long and short time memory network and the full connection layer to obtain time sequence hidden feature vectors; performing attention mechanism processing on the time sequence hidden feature vectors through the attention mechanism layer to generate corresponding attention mechanism feature vectors, performing control parameter analysis on the attention mechanism feature vectors through the inverse normalization layer, and outputting a system dormancy control parameter combination of each target control analysis network; inputting a system dormancy control parameter combination of each target control analysis network into the second-layer strategy optimization network, and generating an initial control strategy population according to the system dormancy control parameter combination through a genetic algorithm in the second-layer strategy optimization network, wherein the initial control strategy population comprises a plurality of initial system dormancy control strategies; calculating adaptation data of each initial system sleep control strategy respectively, and carrying out propagation, intersection and mutation operations on the plurality of initial system sleep control strategies according to the adaptation data to generate a plurality of candidate system sleep control strategies; and optimally selecting the plurality of candidate system sleep control strategies to generate a target system sleep control strategy corresponding to the BMS battery management system.
2. The sleep control method of a BMS battery management system according to claim 1, wherein the acquiring historical electricity consumption data of a target battery pack by a preset BMS battery management system and performing battery load calculation on the historical electricity consumption data to obtain historical electricity consumption load data comprises:
acquiring historical electricity utilization data of a target battery pack through a preset BMS battery management system;
acquiring a voltage attribute tag and a current attribute tag, performing cluster center analysis on the voltage attribute tag through a preset data classification model to determine a corresponding voltage cluster center, and performing cluster center analysis on the current attribute tag through the data classification model to determine a corresponding current cluster center;
inputting the historical electricity consumption data into the data classification model, respectively calculating the data point distances between the historical electricity consumption data and the voltage clustering center and the current clustering center, and generating a corresponding target clustering result according to the data point distances;
generating corresponding historical voltage data and historical current data according to the target clustering result, and performing time stamp alignment on the historical voltage data and the historical current data;
And calculating the product of the historical voltage data and the historical current data to obtain historical electricity load data.
3. The sleep control method of a BMS battery management system according to claim 1, wherein the performing charge-discharge cycle extraction on the historical power consumption load data to obtain a plurality of power consumption cycle load data, and performing curve fitting on the plurality of power consumption cycle load data to generate a plurality of power consumption cycle load curves comprises:
setting a first charging threshold and a first discharging threshold of the target battery pack, and performing charge-discharge state switching detection on the historical electricity load data according to the first charging threshold and the first discharging threshold to obtain a first charge-discharge state switching detection result;
based on the first charge-discharge state switching detection result, carrying out charge-discharge period data set division on the historical electricity load data to obtain a plurality of electricity period load data;
and acquiring first time stamp data corresponding to each power utilization period load data, and respectively performing curve fitting on the power utilization period load data by adopting a first spline interpolation function based on the first time stamp data to generate a plurality of power utilization period load curves.
4. The sleep control method of a BMS battery management system according to claim 3, wherein the obtaining historical system power consumption data of the BMS battery management system, extracting charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and performing curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves, comprises:
acquiring historical system power consumption data of the BMS battery management system, and setting a second charging threshold and a second discharging threshold of the BMS battery management system;
performing charge-discharge state switching detection on the historical system power consumption data according to the second charge threshold and the second discharge threshold to obtain a second charge-discharge state switching detection result;
according to the second charge-discharge state switching detection result, carrying out charge-discharge period data set division on the historical system power consumption data to obtain a plurality of system period power consumption data;
and acquiring second time stamp data corresponding to the power consumption data of each system period, and respectively performing curve fitting on the power consumption data of the plurality of system periods by adopting a second spline interpolation function based on the second time stamp data to generate a plurality of power consumption curves of the system period.
5. The sleep control method of a BMS battery management system according to claim 1, wherein the performing periodic feature extraction on the plurality of power consumption periodic load curves to obtain a plurality of power consumption load feature sets, and performing periodic feature extraction on the plurality of system periodic power consumption curves to obtain a plurality of system power consumption feature sets, respectively, includes:
defining a first periodic characteristic of an electricity periodic load curve, wherein the first periodic characteristic comprises a peak value, a valley value, an average value and a waveform factor;
based on the first periodic characteristics, curve characteristic points are extracted from the plurality of power consumption periodic load curves respectively, so that a plurality of first load curve characteristic points of each power consumption periodic load curve are obtained;
feature screening is carried out on the first load curve feature points to obtain a plurality of second load curve feature points of each power utilization period load curve, a plurality of corresponding power utilization load feature sets are generated according to the second load curve feature points of each power utilization period load curve, and the power utilization load feature sets are in one-to-one correspondence with the power utilization period load curves;
defining a second periodic characteristic of a system periodic power consumption curve, the second periodic characteristic comprising: average value, peak-valley difference, and power consumption fluctuation;
Based on the second periodic characteristics, extracting curve characteristic points of the system periodic power consumption curves respectively to obtain a plurality of first power consumption curve characteristic points of each system periodic power consumption curve;
and performing feature screening on the plurality of first power consumption curve feature points to obtain a plurality of second power consumption curve feature points of each system cycle power consumption curve, and generating a plurality of corresponding system power consumption feature sets according to the plurality of second power consumption curve feature points of each system cycle power consumption curve, wherein the system power consumption feature sets are in one-to-one correspondence with the system cycle power consumption curves.
6. The method for sleep control of a BMS battery management system according to claim 4, wherein the matching the plurality of power load feature sets and the plurality of system power consumption feature sets to obtain a plurality of target fusion feature sets of the same period, and performing vector conversion on the target fusion feature sets to generate a plurality of target period feature vectors, comprises:
matching the plurality of power consumption characteristic sets and the plurality of system power consumption characteristic sets based on the first timestamp data and the second timestamp data to obtain the power consumption characteristic sets and the system power consumption characteristic sets of the same period;
Performing set fusion on the power consumption load feature set and the system power consumption feature set in the same period to generate a plurality of target fusion feature sets in the same period;
and carrying out serialization processing on the target fusion feature set to obtain a target fusion feature sequence, and carrying out vector conversion on the target fusion feature sequence to generate a plurality of target periodic feature vectors.
7. A sleep control device of a BMS battery management system, the sleep control device of the BMS battery management system comprising:
the acquisition module is used for acquiring historical electricity utilization data of the target battery pack through a preset BMS battery management system, and carrying out battery load calculation on the historical electricity utilization data to obtain historical electricity utilization load data;
the fitting module is used for extracting charge and discharge cycles of the historical power utilization load data to obtain a plurality of power utilization cycle load data, and curve fitting the power utilization cycle load data to generate a plurality of power utilization cycle load curves;
the conversion module is used for acquiring historical system power consumption data of the BMS battery management system, extracting charge and discharge cycles of the historical system power consumption data to obtain a plurality of system cycle power consumption data, and performing curve conversion on the plurality of system cycle power consumption data to generate a plurality of system cycle power consumption curves;
The extraction module is used for extracting the periodic characteristics of the plurality of power consumption periodic load curves to obtain a plurality of power consumption load characteristic sets, and extracting the periodic characteristics of the plurality of system periodic power consumption curves to obtain a plurality of system power consumption characteristic sets;
the matching module is used for matching the plurality of power load feature sets with the plurality of system power consumption feature sets to obtain a plurality of target fusion feature sets with the same period, and carrying out vector conversion on the target fusion feature sets to generate a plurality of target period feature vectors;
the analysis module is used for inputting the plurality of target period feature vectors into a preset system dormancy control analysis model to carry out system dormancy control analysis so as to obtain a target system dormancy control strategy; the method specifically comprises the following steps: inputting the target period feature vectors into a preset system dormancy control analysis model, wherein the system dormancy control analysis model comprises a first layer control analysis network and a second layer strategy optimization network, the first layer control analysis network comprises a plurality of target control analysis networks, and each target control analysis network comprises a bidirectional long and short time memory network, a full connection layer, an attention mechanism layer and an inverse normalization layer; receiving the target period feature vectors through a plurality of target control analysis networks in the first layer control analysis network respectively, and performing time sequence period feature operation on the target period feature vectors through the two-way long and short time memory network and the full connection layer to obtain time sequence hidden feature vectors; performing attention mechanism processing on the time sequence hidden feature vectors through the attention mechanism layer to generate corresponding attention mechanism feature vectors, performing control parameter analysis on the attention mechanism feature vectors through the inverse normalization layer, and outputting a system dormancy control parameter combination of each target control analysis network; inputting a system dormancy control parameter combination of each target control analysis network into the second-layer strategy optimization network, and generating an initial control strategy population according to the system dormancy control parameter combination through a genetic algorithm in the second-layer strategy optimization network, wherein the initial control strategy population comprises a plurality of initial system dormancy control strategies; calculating adaptation data of each initial system sleep control strategy respectively, and carrying out propagation, intersection and mutation operations on the plurality of initial system sleep control strategies according to the adaptation data to generate a plurality of candidate system sleep control strategies; and optimally selecting the plurality of candidate system sleep control strategies to generate a target system sleep control strategy corresponding to the BMS battery management system.
8. A sleep control apparatus of a BMS battery management system, the sleep control apparatus of the BMS battery management system comprising: a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invokes the instructions in the memory to cause the sleep control device of the BMS battery management system to perform the sleep control method of the BMS battery management system of any of claims 1-6.
9. A computer readable storage medium having instructions stored thereon, wherein the instructions when executed by a processor implement the sleep control method of the BMS battery management system according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311399178.3A CN117134470B (en) | 2023-10-26 | 2023-10-26 | Sleep control method and related device of BMS battery management system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311399178.3A CN117134470B (en) | 2023-10-26 | 2023-10-26 | Sleep control method and related device of BMS battery management system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117134470A CN117134470A (en) | 2023-11-28 |
| CN117134470B true CN117134470B (en) | 2024-02-13 |
Family
ID=88854961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311399178.3A Active CN117134470B (en) | 2023-10-26 | 2023-10-26 | Sleep control method and related device of BMS battery management system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117134470B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110266070A (en) * | 2019-06-21 | 2019-09-20 | 中国神华能源股份有限公司 | Low-power consumption battery management system, method and railway freight-car |
| CN113780679A (en) * | 2021-09-27 | 2021-12-10 | 深圳供电局有限公司 | Load prediction method and device based on ubiquitous power Internet of things |
| WO2023283846A1 (en) * | 2021-07-14 | 2023-01-19 | 宁德时代新能源科技股份有限公司 | Bms wake-up method and apparatus, and storage medium |
| CN116587921A (en) * | 2023-05-12 | 2023-08-15 | 宁德时代新能源科技股份有限公司 | Battery control method, device, battery management system and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9692240B2 (en) * | 2014-04-30 | 2017-06-27 | Johnson Controls Technology Company | Battery sleep mode management method and system |
-
2023
- 2023-10-26 CN CN202311399178.3A patent/CN117134470B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110266070A (en) * | 2019-06-21 | 2019-09-20 | 中国神华能源股份有限公司 | Low-power consumption battery management system, method and railway freight-car |
| WO2023283846A1 (en) * | 2021-07-14 | 2023-01-19 | 宁德时代新能源科技股份有限公司 | Bms wake-up method and apparatus, and storage medium |
| CN113780679A (en) * | 2021-09-27 | 2021-12-10 | 深圳供电局有限公司 | Load prediction method and device based on ubiquitous power Internet of things |
| CN116587921A (en) * | 2023-05-12 | 2023-08-15 | 宁德时代新能源科技股份有限公司 | Battery control method, device, battery management system and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 锂电池管理系统BMS硬件保护系统架构设计实现;尤勇 等;集成电路应用;第35卷(第08期);第63-66页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117134470A (en) | 2023-11-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| LU502515B1 (en) | Remote cluster charging control method, apparatus and system for charging piles | |
| CN109470957B (en) | Non-invasive electrical appliance identification method based on Frey's distance | |
| CN114720764B (en) | Harmonic analysis method and system based on real-time monitoring data of electric meter | |
| CN110659693A (en) | K-nearest neighbor classification-based rapid topology identification method and system for power distribution network and readable storage medium | |
| EP3870475A1 (en) | Detection of electric vehicle charging at premise | |
| CN116643178B (en) | SOC estimation method and related device of battery management system | |
| CN110901446A (en) | Electric vehicle charging identification method and device based on intelligent electric meter | |
| Liang et al. | Develop load shape dictionary through efficient clustering based on elastic dissimilarity measure | |
| CN113887912A (en) | Non-invasive load identification method for deeply learning downward embedded equipment | |
| CN117134470B (en) | Sleep control method and related device of BMS battery management system | |
| Han et al. | Non‐intrusive load monitoring by voltage–current trajectory enabled asymmetric deep supervised hashing | |
| CN113505465A (en) | Fully unsupervised non-invasive electrical appliance state model self-adaptive construction method | |
| CN117559443A (en) | Ordered power utilization control method for large industrial user cluster under peak load | |
| CN113036754A (en) | Non-invasive load identification method based on improved DAG-SVMS | |
| Afzalan et al. | Semantic search in household energy consumption segmentation through descriptive characterization | |
| CN111506636A (en) | System and method for analyzing residential electricity consumption behavior based on autoregressive and neighbor algorithm | |
| CN105699807A (en) | A supercapacitor electrical characteristic parameter monitoring system and an identification method | |
| CN114662576A (en) | Non-invasive storage battery car charging detection method and system based on supervised classification | |
| Wang et al. | Identifying household EV models via weighted power recurrence graphs | |
| CN112948742A (en) | Non-household load identification method based on average load and total demand distortion rate | |
| CN111815022A (en) | Power load prediction method based on time-delay coordinate embedding method | |
| CN111768066A (en) | Park electric heating load coupling relation analysis method and device based on fusion characteristics | |
| Momtazpour et al. | An integrated data mining framework for analysis and prediction of battery characteristics | |
| CN116861318B (en) | User electricity load classification method, device, equipment and storage medium | |
| CN116512968B (en) | Charging power distribution method, device and equipment based on battery changing cabinet and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |