Distributed energy management intelligent coordination system based on big data

Abstract

The present invention discloses a distributed energy management intelligent coordination system based on big data, including a first acquisition module, a second acquisition module, a data model, an execution module and an energy storage module. The first acquisition module is used to collect the electricity consumption, electricity consumption time and electricity load of the electricity user, and the second acquisition module is used to collect the light intensity e1, light time e2, temperature information f, wind intensity g1, wind direction angle value g2 and dust volume h of the power generator. The information collected by the first acquisition module and the second acquisition module is uploaded to the data model by a data terminal. The present invention makes judgments based on the features extracted from the big data model training, and the auxiliary execution module predicts the electricity consumption behavior of each independent electricity user. At the same time, the execution module can also predict the steep increase and steep decrease of photovoltaic power ΔE and wind power ΔG in advance, so as to facilitate the use of the energy storage module to intervene and coordinate in advance to ensure the stability of the entire system power grid.

Description

Technical Field

[0001] The present invention relates to the field of new energy technology, and in particular to a big data distributed energy management intelligent coordination system. Background Art

[0002] A distributed energy management system refers to a local power grid constructed by multiple groups of small power generation systems. Its power generation system and load constitute an isolated network system that can operate independently. Through energy conversion technology and intelligent control, it achieves efficient use of energy and sustainable development. Compared with the networked power system, the stability of the island power grid has certain defects due to various frequency regulation restrictions and insufficient reactive power compensation control methods.

[0003] The output of renewable energy sources such as solar energy and wind energy is usually intermittent and affected by weather and environmental conditions. This may lead to instability in energy supply, requiring effective energy storage technology or backup power to balance supply and demand. In addition, the user's electricity demand is constantly changing. If the entire distributed energy management system cannot respond to the sudden increase in power load in a timely manner, it will not be able to stabilize the voltage, frequency and phase in the power grid system, causing irreversible damage to the load electrical appliances in the power grid. Therefore, it is a technical problem that technicians in this field need to solve to predict the power demand of the power user and the power generation of the power generator in advance, use energy storage devices to coordinate intelligently, and improve the stability of the distributed energy management system. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method, device and readable storage medium for automatically detecting and rejecting defective products, which solve the problems raised in the above-mentioned background technology.

[0005] The present invention provides a distributed energy management intelligent coordination system based on big data, including a first acquisition module, a second acquisition module, a data model, an execution module and an energy storage module;

[0006] The first acquisition module is used to collect the power consumption, power consumption duration and power load of the power user, and the second acquisition module is used to collect the light intensity e of the power generator. 1 , light exposure time 2 , temperature information f, wind strength g 1 , wind direction angle value g 2 and dust amount h, the information collected by the first acquisition module and the second acquisition module is uploaded to the data model by the data terminal;

[0007] The data model also includes a database, in which all the characteristics of the electricity users and the characteristics of the power generators are recorded. The data model compares the electricity usage duration and the electricity load of the electricity users with the characteristics of the electricity users in the database to analyze the load type in the power grid of the electricity users. The load type is divided into two results: long-term low load and short-term high load. The data model compares the electricity consumption and the electricity usage duration of the electricity users with the characteristics of the power generators in the database to analyze the user groups and user behaviors of the electricity users.

[0008] The data model is based on the light intensity e of the power generation party 1 , light exposure time 2 , dust volume h and temperature information f to analyze the photovoltaic power ΔE of the future photovoltaic power generation. The data model is based on the wind strength g of the power generation party. 1 and wind direction angle g 2 Analyze the wind power ΔG of future wind power generation;

[0009] The data model transmits the analyzed load type, user group, user behavior, photovoltaic power ΔE and wind power ΔG information to the execution module, the execution module calculates the future power load size of the power user according to the load type, user group and user behavior, the execution module calculates the future active power generation of the power generator according to the photovoltaic power ΔE and wind power ΔG, the execution module coordinates the power using the energy storage module according to the supply and demand difference between the power load size and the active power generation, and the execution module uploads the stability result of the coordinated power to the data model for feedback;

[0010] The energy storage module includes a constant micro-charge and discharge module and a sudden charge and discharge module. The execution module predicts that the active power generation is insufficient to meet the power demand of the power load. The execution module will control the electric energy in the constant micro-charge and discharge module to be transmitted to the sudden charge and discharge module in advance to store energy for it, and then release the electric energy through the sudden charge and discharge module to fill the supply and demand gap between the power load size and the active power generation.

[0011] Furthermore, the data information of the power consumption, power consumption duration and power load of the power user is obtained from the user's power meter, and the light intensity e of the power generator is 1 and photoperiod 2 The temperature information f of the power generator is obtained from the ambient light sensor, and the wind strength g of the power generator is obtained from the temperature sensor. 1 and wind direction angle g 2 The information is obtained from the wind speed sensor, and the dust amount h information of the power generation party is obtained from the PM2.5 detector.

[0012] Further, the load type is obtained by the following steps:

[0013] The S01 data model uses the electricity duration as the horizontal axis and the electricity load as the vertical axis to draw a load curve graph, from which the local eigenvalue K is extracted. The calculation formula of the eigenvalue K is as follows:

[0014]

[0015] Among them, x i is the i-th data point in the local area of ​​the curve graph, and n is the number of data points in the curve graph;

[0016] S02 performs model training on the obtained characteristic value K and the characteristics of the electricity users in the database;

[0017] S03 performs preliminary analysis and screening on the obtained invalid features to form a training set and a test set;

[0018] S04 uses the leave-one-out cross-validation method to determine the obtained results. If the error is large, execute S02. Otherwise, if the error is small, execute S05.

[0019] S05 data model outputs the load type and transmits the output result to the execution module.

[0020] Furthermore, the user groups and user behaviors are derived by the following steps:

[0021] S06 data model obtains the original data of power consumption and power consumption duration, and the original data is the specific value of power consumption and power consumption duration quantified in seconds, and the power consumption and power consumption duration are q respectively. t-1 and x t ;

[0022] The S07 data model uses a recurrent neural network to decompose the original data into a data matrix for training. The matrix training process is represented by the following formula group:

[0023]

[0024] Among them, σ is the activation function, which is used to perform nonlinear transformation on the input of neurons, thereby increasing the expressiveness and flexibility of the neural network, and W q , U q and W o is the weight matrix; b q and b o is the bias term, q t is the hidden layer state at the next moment. In the recurrent neural network, the hidden layer consists of multiple neurons and is located between the input layer and the output layer. Its function is to extract and transform the input data in a more complex way, thereby increasing the expression and modeling capabilities of the neural network. t Through the weight matrix W q and U q and the bias term bq Perform linear combination and activation function transformation, output item y t The weight matrix W o and the next hidden layer state q t The linear combination of plus the bias term b o obtained;

[0025] S08 training results in y t The data set is judged using K-fold cross validation. When the error is large, execute S07. Otherwise, if the error result is small, execute S09.

[0026] S09 data model outputs usage groups and user behaviors, and transmits the output results to the execution module.

[0027] Furthermore, the photovoltaic power ΔE is calculated by the following formula:

[0028]

[0029] Among them, a 1 ,a 2 ,a 3 ...,a 7 is the regression coefficient, which can be estimated by fitting the data. The result obtained by this formula is the future photovoltaic power generation predicted by the data model.

[0030] Furthermore, the wind power ΔG is calculated by the following formula:

[0031] Said

[0032] Among them, p 1 ,p 2 ,p 3 ,...,p 5 is the weight factor in the model, and the result obtained by this formula is the future wind power generation predicted by the data model.

[0033] Furthermore, the constant micro charge and discharge module is a large-capacity low-rate battery pack, and the constant micro charge and discharge module is used to support the load type of the power user with long-term low load. The sudden charge and discharge module is a small-capacity high-rate battery pack, and the sudden charge and discharge module is used to support the load type of short-term high load.

[0034] Furthermore, when the active power generated is higher than the power load, the execution module transfers excess energy of the active power generated to the energy storage module, wherein, if any one of the photovoltaic power ΔE and the wind power ΔG predicted by the data model is greater than the charge and discharge power of the constant micro charge and discharge module, the execution module transfers excess energy of the active power generated to the sudden charge and discharge module; conversely, if any one of the predicted photovoltaic power ΔE and the wind power ΔG is less than or equal to the charge and discharge power of the constant micro charge and discharge module, the execution module transfers excess energy of the active power generated to the constant micro charge and discharge module; when the active power generated is lower than the power load, the execution module transfers the electric energy in the storage module to the power user.

[0035] Furthermore, the execution module transmits the excess energy of the active power generated by the power generation to the energy storage module in the following process: the execution module can further refine the charging logic of the energy storage module according to the proportion of the active power generated by the power generation; when the instantaneous power of the excess active power generated by the power generation is lower than or equal to the charge and discharge rate of the large-capacity low-rate battery group, the execution module charges each independent battery in the large-capacity low-rate battery group in turn, and the charging range is 25%-75%. When one of the batteries is charged to 75%, the next battery is charged until all the large-capacity low-rate battery groups are fully charged to 75%. For every 5% increase in the portion of the constant micro-charge and discharge module that exceeds 75%, the execution module limits the charging power of the constant micro-charge and discharge module by 30% until the constant micro-charge and discharge module is fully charged. After the limitation, the execution module transfers the excess electric energy of the active power generated by the power generation to the sudden charge and discharge module; the power output of the active power generated by the power generation is higher than the power load size and can meet the charging and discharging requirements of the constant micro-charge and discharge module. After the power demand is met, the execution module transfers the part of the electric energy that the active power of power generation exceeds instantaneously to the sudden charge and discharge module. If one of the small-capacity high-rate batteries in the small-capacity high-rate battery group is not enough to accommodate the part of the electric energy that the active power of power generation exceeds instantaneously, the execution module controls multiple small-capacity high-rate batteries to charge simultaneously. The charging range of each small-capacity high-rate battery in the sudden charge and discharge module is 25%-75%, which is used to absorb the part of the electric energy that the active power of power generation exceeds instantaneously. Then, when the instantaneous power of the active power of power generation is lower than the charging speed of the constant micro-charge and discharge module, the small-capacity high-rate battery group charges the part of the electric energy that exceeds instantaneously absorbed to another independent large-capacity low-rate battery in the large-capacity low-rate battery group, and the large-capacity low-rate battery corresponding to the charging of the active power of power generation is separated and does not interfere with each other. After the constant micro-charge and discharge module is fully charged, the execution module charges the electric energy of the active power of power generation that is higher than the power load size into the sudden charge and discharge module until it is full.

[0036] The execution module can further refine the discharge logic of the energy storage module according to the proportion of the power load input; when the instantaneous load left after the power load consumes the active power generated is equal to or less than the charge and discharge rate of the large-capacity low-rate battery group, the execution module controls a single large-capacity low-rate battery to discharge the load, and when the continuous discharge of a single large-capacity low-rate battery is less than 25%, it is switched to another independent large-capacity low-rate battery for discharge until the power of all large-capacity low-rate batteries is less than 25%, the execution module preferentially controls the sudden charge and discharge module to discharge until all the power of the sudden charge and discharge module is less than 20%, and the execution module controls all large-capacity low-rate batteries to discharge simultaneously using the remaining 25% of power until all the power is consumed; when the instantaneous load left after the power load consumes the active power generated is greater than the charge and discharge rate of the large-capacity low-rate battery group, the small-capacity high-rate battery A single independent small-capacity high-rate battery in the group is discharged. When a single small-capacity high-rate battery is not enough to meet the extra instantaneous load, the next small-capacity high-rate battery is turned on in turn until all the small-capacity high-rate batteries are turned on for discharge. When the discharge of the small-capacity high-rate battery group still cannot meet the extra instantaneous load, the execution module controls each independent large-capacity low-rate battery in all constant micro-charge and discharge modules to discharge at the same time. When the overall power of the constant micro-charge and discharge module is lower than 25% and the overall power of the sudden charge and discharge module is lower than 20%, the execution module prioritizes limiting the power of the sudden charge and discharge module. For every 5% reduction in the overall power, the current output power is limited to 50%. When the internal power of the sudden charge and discharge module is lower than 10%, the execution module limits the output power of the constant micro-charge and discharge module again. For every 3% reduction in the overall power, the current output power is limited to 33%, until all the electrical energy in the energy storage module is exhausted.

[0037] The present invention has the following beneficial effects:

[0038] 1. The distributed energy management intelligent coordination system based on big data makes judgments through the features extracted from the big data model training, and the auxiliary execution module predicts the electricity consumption behavior of each independent electricity user. At the same time, the execution module can also predict the sharp increase and decrease of photovoltaic power ΔE and wind power ΔG in advance, so as to use the energy storage module to intervene and coordinate in advance to ensure the stability of the entire system power grid.

[0039] 2. The distributed energy management intelligent coordination system based on big data can protect the health of different types of batteries inside the energy storage module as much as possible under complex charging and discharging environments by flexibly calling the energy of the energy storage module through the execution module, thereby extending the service life of the energy storage module.

[0040] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0042] Figure 1 This is a system block diagram of the distributed energy management intelligent coordination system based on big data of the present invention;

[0043] Figure 2 Obtain flow chart for load type;

[0044] Figure 3 Obtain a flow chart for user groups and user behavior. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1-Figure 3 As shown, the present invention provides a technical solution: a distributed energy management intelligent coordination system based on big data, including a first acquisition module, a second acquisition module, a data model, an execution module and an energy storage module;

[0047] The first acquisition module collects the power consumption, power consumption duration and power load of the power user, and the second acquisition module is used to collect the light intensity e of the power generator. 1 , light exposure time 2 , temperature information f, wind strength g 1 , wind direction angle value g 2 and dust amount h, the information collected by the first acquisition module and the second acquisition module is uploaded to the data model by the data terminal. In this embodiment, the specific model of the data terminal is not limited, and any device that can meet the data wireless transmission function in the prior art can be used;

[0048] The data model also includes a database, which records all the characteristics of electricity users and power generators. The data model compares the electricity consumption time and electricity load of the electricity users with the characteristics of the electricity users in the database to analyze the load type in the electricity user's power grid. The load type is divided into two results: long-term low load and short-term high load. The data model compares the electricity consumption and electricity consumption time of the electricity users with the characteristics of the power generators in the database to analyze the user groups and user behaviors of the electricity users.

[0049] The data model is based on the light intensity e of the power generation party. 1 , light exposure time 2 , dust volume h and temperature information f to analyze the photovoltaic power ΔE of the future photovoltaic power generation. The data model is based on the wind strength g of the power generation party. 1 and wind direction angle g 2 Analyze the wind power ΔG of future wind power generation;

[0050] The data model transmits the analyzed load type, user group, user behavior, photovoltaic power ΔE and wind power ΔG information to the execution module. The execution module calculates the future power load size of the power user according to the load type, user group and user behavior. The execution module calculates the future active power generation of the power generator according to the photovoltaic power ΔE and wind power ΔG. The execution module coordinates the power using the energy storage module according to the supply and demand difference between the power load size and the active power generation. The execution module uploads the stability result of the coordinated power to the data model for feedback.

[0051] Among them, the data model predicts the electricity consumption pattern of the electricity user and the power generation pattern of the power generator through the first acquisition module and the second acquisition module. The execution module uses the storage module to coordinate and control the energy in the system according to the predicted pattern. Of course, the prediction of electricity consumption and power generation patterns is not completely accurate. In the system coordination process, the actual active power generation and the power load size are not completely matched. Therefore, when deviations occur, the execution module will feedback the deviation information to the data model to facilitate the data model to correct the features in the database. With the increasing number of independent electricity users and independent power generators, the deviation between the results of the data model calculation and prediction will become smaller and smaller than the actual reality, and the coordination system will become more and more perfect.

[0052] The energy storage module includes a constant micro-charge and discharge module and a sudden charge and discharge module. The execution module predicts that the active power generation is insufficient to meet the power demand of the power load. The execution module will control the electric energy in the constant micro-charge and discharge module to be transmitted to the sudden charge and discharge module in advance to store energy for it, and then release the electric energy through the sudden charge and discharge module to fill the supply and demand gap between the power load size and the active power generation.

[0053] The data information of the electricity consumption, duration of electricity consumption and load of the electricity user is obtained from the user's electricity meter, and the light intensity of the power generator is obtained from the user's electricity meter. 1 and photoperiod 2 The temperature information f of the power generator is obtained from the temperature sensor installed on the photovoltaic panel at the power generator’s location, and the wind strength g of the power generator is obtained from the ambient light sensor deployed at the power user’s location. 1 and wind direction angle g 2 The information is obtained from the wind speed sensor on the wind turbine at the power generation site, and the dust volume h information of the power generation site is obtained from the PM2.5 detector installed on the photovoltaic panel at the power generation site;

[0054] The load type is obtained by the following steps:

[0055] The S01 data model uses the electricity duration as the horizontal axis and the electricity load as the vertical axis to draw a load curve, from which the local eigenvalue K is extracted. The calculation formula of the eigenvalue K is as follows:

[0056]

[0057] Among them, x i is the i-th data point in the local area of ​​the curve graph, and n is the number of data points in the curve graph;

[0058] S02 performs model training on the obtained characteristic value K and the characteristics of the electricity users in the database;

[0059] S03 performs preliminary analysis and screening on the obtained invalid features to form a training set and a test set;

[0060] S04 uses the leave-one-out cross-validation method to determine the obtained results. If the error is large, execute S02. Otherwise, if the error is small, execute S05.

[0061] S05 data model outputs the load type and transmits the output result to the execution module.

[0062] Among them, the user groups and user behaviors are derived from the following steps:

[0063] The S06 data model obtains the original data of power consumption and power consumption duration. The original data is the specific value of power consumption and power consumption duration quantified in seconds. The power consumption and power consumption duration are q respectively. t-1 and x t ;

[0064] The S07 data model uses a recurrent neural network to decompose the original data into a data matrix for training. The matrix training process is represented by the following formula group:

[0065]

[0066] Among them, σ is the activation function, which is used to perform nonlinear transformation on the input of neurons, thereby increasing the expressiveness and flexibility of the neural network, and W q , U q and W o is the weight matrix; b q and b o is the bias term, q t is the hidden layer state at the next moment. In the recurrent neural network, the hidden layer consists of multiple neurons and is located between the input layer and the output layer. Its function is to extract and transform the input data in a more complex way, thereby increasing the expression and modeling capabilities of the neural network. t Through the weight matrix W q and U q and the bias term b q Perform linear combination and activation function transformation, output item y t The weight matrix W o and the next hidden layer state q t The linear combination of plus the bias term b o obtained;

[0067] S08 training results in y t The data set is judged using K-fold cross validation. When the error is large, execute S07. Otherwise, if the error result is small, execute S09.

[0068] S09 data model outputs usage groups and user behaviors, and transmits the output results to the execution module.

[0069] The user groups can be mainly divided into the elderly, adult males, adult females and children, and the judgment is made through the features extracted from the big data model training. There is a certain correlation between the user group and the user behavior, which is used to assist the execution module to predict the electricity consumption behavior of each independent electricity user.

[0070] Through the calculation of the above steps S01-S09, the power consumption pattern of each independent power user in the system can be analyzed by using the model training of big data. The system can predict in advance that a certain power user will use short-term high-load electrical appliances at a specific time. The execution module can transfer the electric energy in the constant micro-charge and discharge module to the sudden charge and discharge module for storage in advance, and release it when the power load increases sharply, thereby ensuring the stability of the entire power system.

[0071] The photovoltaic power ΔE is calculated by the following formula:

[0072]

[0073] Among them, a 1 ,a 2 ,a3 ...,a 7 is the regression coefficient, which can be estimated by fitting data. The result obtained by this formula is the future photovoltaic power generation predicted by the data model. It should be noted that the temperature information f involved in the formula is mainly used to express the light conversion efficiency on the photovoltaic panel. High temperature will reduce the electron mobility of the semiconductor material in the panel, thereby reducing the output power of the panel. High temperature may also cause aging and damage to the panel, shortening its service life. In addition, the amount of dust h is also an important indicator affecting the photovoltaic power ΔE. Dust will reduce the transmittance of the solar panel and reduce the amount of incident light, thereby affecting the power generation efficiency of the solar panel. If dust accumulates for a long time, it may also form dirt on the surface of the solar panel, further hindering the transmission of light. This will not only reduce the power generation efficiency, but may also cause local overheating, which will have an adverse effect on the life of the solar panel.

[0074] The wind power ΔG is calculated by the following formula:

[0075]

[0076] Among them, p 1 ,p 2 ,p 3 ,...,p 5 is the weight factor in the model, and the result obtained by this formula is the future wind power generation predicted by the data model.

[0077] Through the above calculation method, the execution module in the system can predict the sharp increase and decrease of photovoltaic power ΔE and wind power ΔG in advance, which is convenient for using energy storage modules to intervene and coordinate to ensure the stability of the entire system power grid.

[0078] Among them, the constant micro charge and discharge module is a large-capacity, low-rate battery group. When the active power generation is insufficient, the constant micro charge and discharge module is used to support the long-term low-load load type of the electricity user. The sudden charge and discharge module is a small-capacity, high-rate battery group. When the active power generation is insufficient, the sudden charge and discharge module is used to support the short-term high-load load type.

[0079] When the active power of power generation is higher than the power load, the execution module transmits the excess energy of the active power of power generation to the energy storage module. If any one of the photovoltaic power ΔE and wind power ΔG predicted by the data model is greater than the charge and discharge power of the constant micro-charge and discharge module, the execution module transmits the excess energy of the active power of power generation to the sudden charge and discharge module. On the contrary, if any one of the predicted photovoltaic power ΔE and wind power ΔG is less than or equal to the charge and discharge power of the constant micro-charge and discharge module, the execution module transmits the excess energy of the active power of power generation to the constant micro-charge and discharge module.

[0080] Among them, the specific process of the execution module transmitting the excess energy of the active power of power generation to the energy storage module is as follows: according to the proportion of the active power output of power generation, the execution module can further refine the charging logic of the energy storage module; when the instantaneous power of the excess active power of power generation is lower than or equal to the charge and discharge rate of the large-capacity low-rate battery group, the execution module charges each independent battery in the large-capacity low-rate battery group in turn, and the charging range is 25%-75%. When one of the batteries is charged to 75%, it is at the energy saturation warning value of the large-capacity low-rate battery, and the next battery is charged until all large-capacity low-rate battery groups are fully charged to 75%. For every 5% increase in the charging of the constant micro-charge and discharge module exceeding 75%, the execution module limits the charging power of the constant micro-charge and discharge module by 30% until the constant micro-charge and discharge module is fully charged, which can further protect the health of the batteries inside the constant micro-charge and discharge module, and transfer the excess electric energy of the active power of power generation to the mutation charge and discharge module after the restriction; the power output of the active power of power generation is higher than the power load size and can meet the constant micro-charge After the discharge module meets the charging demand, the execution module transfers the part of the electric energy that exceeds the active power generation instantaneously to the mutation charge and discharge module. If one of the small-capacity high-rate batteries in the small-capacity high-rate battery group is not enough to accommodate the part of the electric energy that exceeds the active power generation instantaneously, the execution module controls multiple small-capacity high-rate batteries to charge at the same time. The charging range of each small-capacity high-rate battery in the mutation charge and discharge module is 25%-75%, which is used to absorb the part of the electric energy that exceeds the active power generation instantaneously. Then, when the instantaneous power of the active power generation is lower than After the constant micro charge and discharge module is charged, the small capacity high rate battery pack will absorb part of the instantaneous excess electric energy to another independent large capacity low rate battery in the large capacity low rate battery pack, and the large capacity low rate battery corresponding to the active power charging will be separated and will not interfere with each other, giving priority to the large capacity low rate battery pack to store electric energy, while ensuring the stability of the low rate battery pack. After the constant micro charge and discharge module is fully charged, the execution module will charge the electric energy with the active power higher than the power load into the sudden charge and discharge module until it is full;

[0081] When the active power generated is less than or equal to the power load, the execution module transmits the electric energy in the storage module to the power user; specifically:

[0082] The execution module can further refine the discharge logic of the energy storage module according to the proportion of the power load input; when the instantaneous load left after the power load consumes the active power generated is equal to or less than the charge and discharge rate of the large-capacity low-rate battery group, the execution module controls a single large-capacity low-rate battery to discharge the load. When the continuous discharge of a single large-capacity low-rate battery is less than 25%, it is switched to another independent large-capacity low-rate battery for discharge until the power of all large-capacity low-rate batteries is less than 25%. The execution module preferentially controls the sudden charge and discharge module to discharge until all the power of the sudden charge and discharge module is less than 20%. When the sudden charge and discharge module is at a low power warning value, the execution module controls all large-capacity low-rate batteries to discharge simultaneously using the remaining 25% power until all the power is consumed. When the instantaneous load left after the power load consumes the active power generated is greater than the charge and discharge rate of the large-capacity low-rate battery group, a single independent small-capacity high-rate battery in the small-capacity high-rate battery group is When a single small-capacity high-rate battery is not enough to meet the extra instantaneous load, the next small-capacity high-rate battery is turned on in turn until all the small-capacity high-rate batteries are turned on for discharge to ensure that the steeply rising instantaneous load is met. When the discharge of the small-capacity high-rate battery group still cannot meet the extra instantaneous load, the execution module controls each independent large-capacity low-rate battery in all constant micro-charge and discharge modules to discharge at the same time. When the overall power of the constant micro-charge and discharge module is lower than 25% and the overall power of the sudden charge and discharge module is lower than 20%, the execution module prioritizes limiting the power of the sudden charge and discharge module. For every 5% reduction in the overall power, the current output power is limited to 50%, to ensure the health of the small-capacity high-rate batteries in the sudden charge and discharge module. When the internal power of the sudden charge and discharge module is lower than 10%, the execution module limits the output power of the constant micro-charge and discharge module again. For every 3% reduction in the overall power, the current output power is limited to 33%, until all the power in the energy storage module is exhausted.

[0083] It should be noted that the various letter numbers involved in the embodiments of the present invention are only distinguished for the convenience of description and are not used to limit the scope of the embodiments of the present application.

[0084] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A distributed energy management intelligent coordination system based on big data, comprising a first acquisition module, a second acquisition module, a data model, an execution module and an energy storage module; The first acquisition module is used to collect the electricity consumption information of the electricity user, and the second acquisition module is used to collect the power generation information of the power generator. The electricity consumption information and the power generation information are uploaded to the data model by the data terminal; The electricity usage information includes electricity usage, electricity usage duration and electricity load, and the power generation information includes light intensity, light duration, temperature information, wind strength, wind direction angle value and dust volume; the data model compares the electricity usage duration and electricity load of the electricity user with the characteristics of the electricity user in the database to analyze the load type in the electricity user's power grid, and the load type is divided into two results: long-term low load and short-term high load. The data model compares the electricity usage and electricity usage duration of the electricity user with the characteristics of the power generation party in the database to analyze the user group and user behavior of the electricity user; The data model analyzes the photovoltaic power ΔE of the future photovoltaic power generation according to the light intensity e1, light time e2, dust amount h and temperature information f of the power generator, and the data model analyzes the wind power ΔG of the future wind power generation according to the wind intensity g1 and wind direction angle value g2 of the power generator; The data model transmits the analyzed load type, user group, user behavior, photovoltaic power ΔE and wind power ΔG information to the execution module, the execution module calculates the future power load size of the power user according to the load type, user group and user behavior, the execution module calculates the future active power generation of the power generator according to the photovoltaic power ΔE and wind power ΔG, the execution module coordinates the power using the energy storage module according to the supply and demand difference between the power load size and the active power generation, and the execution module uploads the stability result of the coordinated power to the data model for feedback; The energy storage module includes a constant micro charge and discharge module and a sudden charge and discharge module; the characteristics are: The load type is derived from the following steps: S01: The data model uses the electricity duration as the horizontal axis and the electricity load as the vertical axis to draw a load curve graph, from which a local eigenvalue K is extracted. The calculation formula of the eigenvalue K is as follows: Among them, x i is the i-th data point in the local area of ​​the curve graph, and n is the number of data points in the curve graph; S02: Perform model training using the obtained eigenvalue K and the electricity user characteristics in the database; S03: Perform preliminary analysis and screening on the obtained invalid features to form training sets and test sets; S04: Use leave-one-out cross validation to determine the result. If the error is large, execute S02. Otherwise, if the error is small, execute S05. S05: The data model outputs the load type and transmits the output result to the execution module; The user groups and user behaviors are derived from the following steps: S06: The data model obtains the original data of power consumption and power consumption duration, where the power consumption and power consumption duration are q t-1 and x t ; S07: The data model uses a recurrent neural network to decompose the original data into a data matrix for training. The matrix training process is represented by the following formula group: Among them, σ is the activation function, which performs nonlinear transformation on the input of the neuron, W q , U q and W o is the weight matrix; b q and b o is the bias term, q t is the state of the hidden layer at the next moment, q t Through the weight matrix W q and U q and the bias term b q Perform linear combination and activation function transformation, output item y t The weight matrix W o and the next hidden layer state q t The linear combination of plus the bias term b o obtained; S08: Training to get y t The data set is judged using K-fold cross validation. When the error is large, execute S07. Otherwise, if the error result is small, execute S09. S09: The data model outputs the user group and user behavior, and transmits the output result to the execution module; The photovoltaic power ΔE is calculated by the following formula: Among them, a1, a2, a3..., a7 are regression coefficients, which are estimated by fitting the data; The wind power ΔG is calculated by the following formula: Among them, p1, p2, p3, ..., p5 are weight factors in the model.

2. According to the big data-based distributed energy management intelligent coordination system of claim 1, it is characterized in that: The data information of the electricity consumption, electricity consumption duration and electricity load of the electricity user is obtained from the user's electricity meter, the light intensity e1 and lighting time e2 of the power generator are obtained from the ambient light sensor, the temperature information f of the power generator is obtained from the temperature sensor, the wind intensity g1 and wind direction angle value g2 of the power generator are obtained from the wind speed sensor, and the dust amount h information of the power generator is obtained from the PM2.5 detector.

3. The distributed energy management intelligent coordination system based on big data according to claim 1 is characterized in that: The constant micro-charge and discharge module is a large-capacity low-rate battery pack, which is used to support the load type of long-term low load of the power user. The sudden charge and discharge module is a small-capacity high-rate battery pack, which is used to support the load type of short-term high load. The execution module balances and coordinates the active power generation and the power load size through the constant micro-charge and discharge module and the sudden charge and discharge module in the energy storage module. The specific process of the execution module balancing and coordinating the active power generation and the power load size is as follows: When the active power generated is higher than the power load, the execution module transmits the excess energy of the active power generated to the energy storage module, wherein if any one of the photovoltaic power ΔE and the wind power ΔG predicted by the data model is greater than the charge and discharge power of the constant micro charge and discharge module, the execution module transmits the excess energy of the active power generated to the mutation charge and discharge module; conversely, if any one of the predicted photovoltaic power ΔE and the wind power ΔG is less than or equal to the charge and discharge power of the constant micro charge and discharge module, the execution module transmits the excess energy of the active power generated to the constant micro charge and discharge module; When the generated active power is lower than the power load, the execution module transmits the electric energy in the storage module to the power user.

Images