A mobile energy storage vehicle coordination scheduling system based on NB-IoT

Abstract

The application discloses a kind of mobile energy storage vehicle coordination scheduling systems based on NB-IoT, it is related to energy storage scheduling technical field, including perception layer, NB-IoT communication layer, edge computing layer and cloud scheduling layer, each layer is sequentially communicated and connected;Perception layer includes battery state monitoring module, position positioning module, power distribution network parameter acquisition module and environmental monitoring module, battery state monitoring module is used to collect the voltage, temperature, state of charge, health status and charge-discharge power parameter of mobile energy storage vehicle battery, position positioning module is used to obtain the real-time position and driving trajectory of mobile energy storage vehicle.The application relies on the low power consumption, wide coverage advantage of NB-IoT technology, combines edge computing and artificial intelligence algorithm, improves scheduling response speed and coordination accuracy, reduces scheduling cost, enhances the resilience of power distribution network to deal with distributed energy fluctuation, and can be widely applied to power distribution network peak shaving, emergency power supply and renewable energy consumption scene.

Description

Technical Field

[0001] This invention relates to the field of energy storage scheduling technology, specifically to a mobile energy storage vehicle collaborative scheduling system based on NB-IoT. Background Technology

[0002] With the continuous increase in the penetration rate of renewable energy sources such as distributed photovoltaic and wind power, the strong randomness and spatiotemporal uncertainty of their output significantly exacerbate the operational risks of distribution networks, leading to local node voltage overruns, frequent reverse power flow, and even protection malfunctions and equipment overloads. Mobile energy storage vehicles, due to their advantages of flexible deployment and rapid dynamic response, have become an effective means to alleviate these problems and improve the resilience of distribution networks, and are widely used in scenarios such as distribution network peak shaving, emergency power supply, and renewable energy consumption.

[0003] NB-IoT (Narrowband Internet of Things), as a low-power wide-area communication technology, boasts advantages such as wide coverage, low power consumption, low cost, and high connection capacity, and has been widely applied in fields such as intelligent monitoring and remote control. Combining NB-IoT technology with mobile energy storage vehicle scheduling can effectively solve the communication bottlenecks of existing systems. Furthermore, integrating edge computing and artificial intelligence algorithms can improve data processing efficiency and collaborative scheduling accuracy, achieving optimal spatiotemporal configuration and dynamic coordination of charging and discharging for mobile energy storage vehicles. However, existing mobile energy storage vehicle scheduling systems still suffer from the following problems: Most communication methods employ traditional wireless communication technologies, which suffer from limited coverage, high power consumption, and significant communication latency. This makes it difficult to achieve real-time data exchange and collaborative scheduling among multiple mobile energy storage vehicles, especially in remote areas or complex terrains where communication stability is difficult to guarantee and collaborative scheduling capabilities are insufficient. Many systems employ an independent scheduling mode for a single energy storage vehicle, failing to fully consider the coordination between multiple vehicles, resulting in low resource utilization and hindering overall power distribution network optimization. To address this, we present an NB-IoT-based collaborative scheduling system for mobile energy storage vehicles. Summary of the Invention

[0004] The purpose of this invention is to provide a mobile energy storage vehicle collaborative scheduling system based on NB-IoT to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A mobile energy storage vehicle collaborative scheduling system based on NB-IoT includes a sensing layer, an NB-IoT communication layer, an edge computing layer, and a cloud scheduling layer, with each layer communicating with each other in sequence; The sensing layer includes a battery status monitoring module, a location positioning module, a power distribution network parameter acquisition module, and an environmental monitoring module. The battery status monitoring module is used to collect the voltage, temperature, state of charge, health status, and charge / discharge power parameters of the mobile energy storage vehicle's battery. The location positioning module is used to obtain the real-time location and driving trajectory of the mobile energy storage vehicle. The power distribution network parameter acquisition module is used to collect the voltage, power, and load demand data of the power distribution network nodes. The environmental monitoring module is used to collect meteorological parameters such as ambient temperature, humidity, and precipitation in the driving path and operating area.

[0006] Preferably, the NB-IoT communication layer includes an NB-IoT terminal module, an NB-IoT base station, and a communication gateway. The NB-IoT terminal module is integrated into each acquisition unit of the sensing layer and the mobile energy storage vehicle control terminal. It is used to encapsulate the acquired data into NB-IoT communication protocol data frames, upload the data to the communication gateway through the NB-IoT base station, and receive scheduling instructions issued by the edge computing layer and the cloud scheduling layer.

[0007] Preferably, the edge computing layer includes a data preprocessing module, a local collaborative decision-making module, a prediction error compensation module, and an instruction issuing module. The data preprocessing module is used to perform noise reduction, normalization, and outlier removal on the data uploaded by the perception layer. The local collaborative decision-making module is used to realize short-term collaborative scheduling of multiple mobile energy storage vehicles within the region. The prediction error compensation module is used to correct the prediction deviations of energy output and load demand. The instruction issuing module is used to issue scheduling instructions to the mobile energy storage vehicles through the NB-IoT communication layer.

[0008] Preferably, the local collaborative decision-making module of the edge computing layer adopts a reinforcement learning algorithm to generate a short-term scheduling plan based on the real-time load of the distribution network and the status of mobile energy storage vehicles in the region; The prediction error compensation module employs scene generation and reduction technology, generating a typical source-load scene set through K-means clustering to correct prediction bias.

[0009] Preferably, the cloud-based scheduling layer includes a data storage module, a multi-objective optimization scheduling module, a status monitoring module, and a human-machine interaction module. The data storage module is used to store historical data, real-time data, and scheduling schemes. The multi-objective optimization scheduling module is used to construct a multi-objective optimization model that includes minimizing voltage deviation, minimizing network loss, minimizing scheduling cost, and maximizing battery life, and uses an improved particle swarm optimization algorithm to solve for the optimal scheduling scheme. The status monitoring module is used to monitor the operating status of the mobile energy storage vehicle and the distribution network in real time. The human-machine interaction module is used to realize the setting of scheduling parameters, the display of scheduling schemes, and abnormal alarms.

[0010] Preferably, it also includes a security protection module, which uses two-way authentication and data encryption technology to encrypt NB-IoT communication data and scheduling instructions to prevent data tampering and unauthorized access; at the same time, it sets up battery overcharge, over-discharge and over-temperature protection mechanisms, and automatically triggers an alarm and cuts off the charging and discharging circuit when abnormal parameters are detected.

[0011] Preferably, the cloud scheduling layer supports a rolling time-domain execution mechanism, which updates the source-load prediction data and mobile energy storage vehicle status data periodically, re-solves the optimization model, extracts and executes the current scheduling instructions, and forms a closed-loop feedback.

[0012] Preferably, the local collaborative decision-making module specifically includes: A Markov decision process is constructed using a deep reinforcement learning algorithm, and its quintuple is defined as follows: in, The state space contains real-time state quantities such as the load of the regional distribution network, the output of distributed energy resources, the SOC and location of mobile energy storage vehicles, and the voltage of distribution network nodes. The action space includes scheduling actions such as adjusting the charging and discharging power and adjusting the driving path of the mobile energy storage vehicle; Here is the state transition probability matrix. Indicates the state Next action Then, the system transitions to state. The probability of; This is a discount factor with a value of 0.95, used to balance immediate rewards with long-term benefits; Here is a single-step reward function used to evaluate the quality of scheduling actions; the formula is: in, For node voltage deviation, For the increase in distribution network losses, The total scheduling cost, These are weighting coefficients, with values ​​of 0.4, 0.3, and 0.3 respectively. The Q-learning algorithm is used for decision-making, and the temporal difference update formula for the Q value is: in, The learning rate is 0.1. For state Next action Action value function; For the next state The maximum value of the action.

[0013] Preferably, the prediction error compensation module specifically includes: Based on historical source load data, the prediction error of load and energy output is calculated: in, , They are respectively Load forecasting error and energy output forecasting error at any given time; , These are the actual values ​​for load and energy output, respectively. , These are the predicted values ​​for load and energy output, respectively, both in kW; Then, based on the probability distribution of the prediction error, a large number of source load error scenarios are generated. Subsequently, the K-means clustering algorithm is used to reduce the number of scenarios to obtain typical error scenarios. The objective function of clustering is: in, The clustering loss function; Number of typical scenarios; For the first A collection of scenarios for each cluster; The source-load error scenario sample vector; For the first The centroid vector of each cluster.

[0014] Preferably, the objective function of the multi-objective optimization scheduling model used by the multi-objective optimization scheduling module is: in, For the scheduling period, , , , These are the weighting coefficients. For nodes exist Voltage amplitude at time 10:00 Rated voltage; For the line exist Current at any moment For the line The resistance, For the first The unit dispatch cost of a mobile energy storage vehicle For the first Mobile energy storage vehicles in The charging and discharging power at any given time For the first Mobile energy storage vehicles in The cost of battery degradation over time.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention uses NB-IoT communication technology to achieve wide coverage, low power consumption, and high reliability data transmission. The communication latency is no more than 100ms and the communication success rate is no less than 99.5%. It solves the problems of insufficient coverage and high latency of traditional communication technologies in remote areas or complex terrains. At the same time, it reduces the energy consumption of mobile energy storage vehicles and improves the system's endurance.

[0016] This invention constructs a two-layer collaborative architecture of edge computing and cloud scheduling. The edge computing layer realizes short-term collaborative scheduling within the region, while the cloud scheduling layer realizes global optimal scheduling. By combining the improved particle swarm optimization algorithm and prediction error compensation technology, it effectively copes with source-load uncertainty, improves the accuracy and adaptability of the scheduling scheme, and realizes the joint optimization of path and power of mobile energy storage vehicles.

[0017] This invention integrates multiple types of high-precision sensors in the sensing layer to achieve comprehensive collection of battery status, power grid parameters, and environmental parameters of mobile energy storage vehicles. Combined with data preprocessing and fusion technologies, it improves the reliability and accuracy of the data, providing precise data support for the formulation of scheduling schemes. At the same time, it enables dynamic assessment of battery health status and extends battery life.

[0018] This invention uses an edge computing layer to process real-time data locally, reducing the computing pressure on the cloud and enabling the rapid generation and issuance of short-term scheduling plans and instructions. The emergency response time is no more than 5 minutes, which can quickly respond to distribution network faults and sudden load peaks, thus improving the resilience of the distribution network.

[0019] This invention establishes a comprehensive security protection mechanism to ensure system data security and equipment operation safety; through multi-objective optimized scheduling, it achieves comprehensive optimization of voltage deviation, network loss, scheduling cost and battery life, reduces the operating cost of the distribution network, and improves the renewable energy absorption capacity, resulting in significant economic and social benefits. Attached Figure Description

[0020] Figure 1 This is a diagram of the overall system architecture of the present invention; Figure 2 This is a schematic diagram of the sensing layer structure of the present invention; Figure 3 This is a flowchart of the closed-loop collaborative scheduling process of the system of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] like Figure 1-3 As shown, a mobile energy storage vehicle collaborative scheduling system based on NB-IoT includes a sensing layer, an NB-IoT communication layer, an edge computing layer, and a cloud scheduling layer. Each layer is connected in sequence to form a closed-loop collaborative scheduling architecture. The sensing layer includes a battery status monitoring module, a location positioning module, a power distribution network parameter acquisition module, and an environmental monitoring module. The battery status monitoring module is used to collect parameters such as voltage, temperature, state of charge (SOC), state of health (SOH), and charge / discharge power of the mobile energy storage vehicle's battery. The battery status monitoring module uses high-precision voltage sensors, temperature sensors, and SOC / SOH detection chips, which are installed in the battery pack of the mobile energy storage vehicle. It collects key parameters such as battery voltage, temperature, SOC, SOH, and charge / discharge power in real time, with a collection frequency of 1-5 minutes / time. The error of the collected data does not exceed ±2%, which can accurately reflect the battery's operating status and health level, providing a basis for battery life protection and charge / discharge scheduling. Among them, the SOC detection uses a Kalman filter algorithm for optimization to improve detection accuracy; the SOH detection realizes dynamic assessment of battery health status by monitoring parameters such as battery capacity decay and internal resistance changes.

[0023] The location positioning module is used to obtain the real-time location and driving trajectory of the mobile energy storage vehicle. The location positioning module adopts GPS + Beidou dual-mode positioning and is integrated into the vehicle terminal of the mobile energy storage vehicle. It obtains the longitude, latitude, driving speed and driving trajectory of the mobile energy storage vehicle in real time with a positioning accuracy of no less than 10 meters. It supports dynamic trajectory real-time updates and can accurately grasp the location information of each mobile energy storage vehicle, providing spatial data support for path planning and collaborative scheduling. At the same time, the module has an offline positioning function. When NB-IoT communication is interrupted, it can store location data and automatically upload it after communication is restored.

[0024] The distribution network parameter acquisition module is used to collect voltage, power, and load demand data of distribution network nodes. The module adopts NB-IoT wireless sensors, which do not require wiring and can be flexibly deployed at various nodes of the distribution network. It collects voltage, current, power, and load demand data of the nodes, supports data interaction with the existing monitoring system of the distribution network, is compatible with the IEC61850 communication protocol, and can monitor the operating status of the distribution network in real time, providing power grid data support for the formulation of dispatching plans.

[0025] The environmental monitoring module is used to collect meteorological parameters such as ambient temperature, humidity, and precipitation along the driving route and in the operating area. The environmental monitoring module uses temperature and humidity sensors, precipitation sensors, and other equipment to collect meteorological parameters such as ambient temperature, humidity, precipitation, and wind speed along the driving route and in the operating area of ​​the mobile energy storage vehicle. This data is used to assess the impact of environmental factors on battery performance and driving safety, and to provide a basis for route planning and adjustment of charging and discharging parameters.

[0026] The NB-IoT communication layer includes an NB-IoT terminal module, an NB-IoT base station, and a communication gateway. The NB-IoT terminal module is integrated into each acquisition unit of the perception layer and the control terminal of the mobile energy storage vehicle. It is used to encapsulate the acquired data into NB-IoT communication protocol data frames, upload the data to the communication gateway through the NB-IoT base station, and receive scheduling instructions issued by the edge computing layer and the cloud scheduling layer. The NB-IoT terminal module is integrated into the various acquisition units of the sensing layer and the control terminal of the mobile energy storage vehicle. It adopts a low-power design, supports the NB-IoT communication protocol, and can encapsulate the acquired data into standard data frames. It can then upload the data through the NB-IoT base station and receive scheduling instructions from the edge computing layer and the cloud scheduling layer. The module supports a sleep-wake function, entering a sleep state when there is no data transmission, thereby reducing the energy consumption of the mobile energy storage vehicle.

[0027] NB-IoT base stations are deployed within the coverage area and are responsible for receiving data uploaded by NB-IoT terminal modules and forwarding it to the communication gateway. At the same time, they forward the instructions issued by the communication gateway to the corresponding NB-IoT terminal modules. The base station adopts narrowband orthogonal frequency division multiplexing technology and the communication frequency is 800-900MHz. It has the advantages of wide coverage and strong anti-interference ability, and can achieve signal coverage in remote areas and complex terrains.

[0028] As a bridge connecting the NB-IoT communication layer with the edge computing layer and the cloud scheduling layer, the communication gateway is responsible for protocol conversion of the data uploaded by the NB-IoT base station, converting the NB-IoT protocol into the Ethernet protocol or TCP / IP protocol to facilitate data processing by the edge computing layer and the cloud scheduling layer. At the same time, it converts the scheduling instructions issued by the edge computing layer and the cloud scheduling layer into the NB-IoT protocol and sends them to the mobile energy storage vehicle through the NB-IoT base station. The communication gateway supports concurrent connections of multiple devices and can handle data transmission and instruction interaction of multiple mobile energy storage vehicles at the same time, with a communication latency of no more than 100ms and a communication success rate of no less than 99.5%.

[0029] The edge computing layer includes a data preprocessing module, a local collaborative decision-making module, a prediction error compensation module, and a command issuance module. The data preprocessing module is used to perform noise reduction, normalization, and outlier removal on the data uploaded from the perception layer. This includes data noise reduction, normalization, outlier removal, and data fusion processing. Wavelet transform algorithm is used for data noise reduction to remove noise caused by environmental interference and equipment errors; Z-score normalization is used for data normalization to facilitate subsequent data processing and algorithm solving; 3σ criterion is used to remove outliers to avoid the impact of abnormal data on the scheduling scheme; and data fusion algorithm integrates data collected from multiple sensors to improve data reliability and integrity.

[0030] The local collaborative decision-making module is used to realize short-term collaborative scheduling of multiple mobile energy storage vehicles within a region. Employing a reinforcement learning algorithm, the module generates a 15-30 minute short-term scheduling plan based on the real-time load of the regional power distribution network, the status of the mobile energy storage vehicles, and environmental parameters. This enables local collaborative scheduling of multiple mobile energy storage vehicles within the region, specifically including: A Markov decision process is constructed using a deep reinforcement learning algorithm, and its quintuple is defined as follows: in, The state space includes real-time state variables such as the load of the regional distribution network, the output of distributed photovoltaic / wind power, the SOC and location of mobile energy storage vehicles, and the voltage of distribution network nodes. The action space includes scheduling actions such as adjusting the charging and discharging power and adjusting the driving path of the mobile energy storage vehicle; Here is the state transition probability matrix. Indicates the state Next action Then, the system transitions to state. The probability of; This is a discount factor with a value of 0.95, used to balance immediate rewards with long-term benefits; Here is a single-step reward function used to evaluate the quality of scheduling actions; the formula is: in, For node voltage deviation, For the increase in distribution network losses, The total scheduling cost, These are weighting coefficients, with values ​​of 0.4, 0.3, and 0.3 respectively. The Q-learning algorithm is used for decision-making, and the temporal difference update formula for the Q value is: in, The learning rate is 0.1. For state Next action Action value function; For the next state The maximum value of the action; Through the above algorithm, the module can adjust the scheduling actions of mobile energy storage vehicles in real time according to the real-time operating status of the area, realize short-term collaborative optimization within the area, and effectively smooth short-term fluctuations in the distribution network.

[0031] The prediction error compensation module is used to correct prediction deviations in photovoltaic / wind power output and load demand. Addressing the randomness and uncertainty of distributed photovoltaic / wind power output and load demand, the module employs scenario generation and reduction techniques. It generates a set of typical source-load scenarios through K-means clustering to correct prediction deviations. Specifically, this includes: Based on historical source-load data, the prediction error between load and photovoltaic / wind power output is calculated: in, , They are respectively Load forecasting error at any given time and photovoltaic / wind power output forecasting error; , These are the actual values ​​for load and photovoltaic / wind power output, respectively. , These are the predicted values ​​for load and photovoltaic / wind power output, respectively, both in kW; Then, based on the probability distribution of the prediction error, a large number of source load error scenarios are generated. Subsequently, the K-means clustering algorithm is used to reduce the number of scenarios to obtain typical error scenarios. The objective function of clustering is: in, The clustering loss function; Number of typical scenarios; For the first A collection of scenarios for each cluster; The source-load error scenario sample vector; For the first The centroid vectors of the clusters; By analyzing typical error scenarios, the module corrects the offline scheduling schemes issued from the cloud, compensates for scheduling deviations caused by prediction errors, and improves the actual operating effect of the scheduling schemes.

[0032] The instruction delivery module is used to send scheduling instructions to the mobile energy storage vehicle through the NB-IoT communication layer. The instruction delivery module sends the local scheduling instructions generated by the edge computing layer and the global scheduling instructions sent by the cloud scheduling layer to the corresponding mobile energy storage vehicle through the NB-IoT communication layer. At the same time, it receives the execution feedback information from the mobile energy storage vehicle to realize closed-loop monitoring of instruction execution.

[0033] The cloud-based scheduling layer comprises a data storage module, a multi-objective optimization scheduling module, a status monitoring module, and a human-computer interaction module. The data storage module stores historical data, real-time data, and scheduling schemes. It utilizes a cloud database to store real-time data uploaded from the perception layer, data processed by the edge computing layer, historical scheduling data, scheduling schemes, and system parameters. The database supports encrypted storage and backup to ensure data security and integrity, while also supporting data querying and export, providing data support for scheduling optimization and system upgrades.

[0034] The multi-objective optimization scheduling module is used to construct a multi-objective optimization model that includes minimizing voltage deviation, minimizing network loss, minimizing scheduling cost, and maximizing battery life. An improved particle swarm optimization algorithm is used to solve for the optimal scheduling scheme, achieving joint path-power optimization and global coordination for mobile energy storage vehicles. The objective function of the multi-objective optimization scheduling model is: in, For the scheduling period, , , , These are the weighting coefficients. For nodes exist Voltage amplitude at time 10:00 Rated voltage; For the line exist Current at any moment For the line The resistance, For the first The unit dispatch cost of a mobile energy storage vehicle For the first Mobile energy storage vehicles in The charging and discharging power at any given time For the first Mobile energy storage vehicles in The cost of battery degradation over time; The multi-objective optimization scheduling module introduces an adaptive adjustment strategy for inertial weights, combined with the global search capability of the simulated annealing algorithm, to avoid the algorithm getting trapped in local optima and improve the solution speed compared to the traditional particle swarm optimization algorithm. Simultaneously, the module supports a rolling time-domain execution mechanism, updating source-load prediction data and mobile energy storage vehicle status data every 15 minutes, resolving the optimization model, extracting and executing the current-moment scheduling instructions, forming a closed-loop feedback, and improving the real-time performance and adaptability of the scheduling.

[0035] The status monitoring module is used to monitor the operation status of the mobile energy storage vehicle and the power distribution network in real time. The status monitoring module monitors the operation status of the mobile energy storage vehicle (including battery status, location, and driving status), the operation status of the power distribution network, and the working status of each layer of the system in real time. When abnormal parameters are detected (such as battery overcharging, over-discharging, over-temperature, power distribution network voltage exceeding the limit, communication interruption, etc.), an alarm is automatically triggered, and the abnormal information is pushed to the human-machine interaction module to facilitate timely handling by staff and ensure the safe and stable operation of the system.

[0036] The human-computer interaction module is used to set scheduling parameters, display scheduling schemes, and generate alarms. It employs a visual interface to provide functions such as setting scheduling parameters, displaying scheduling schemes, viewing alarms, querying data, and generating reports. Staff can use this module to set parameters such as scheduling cycles, weighting coefficients, and safety thresholds, view real-time scheduling schemes and operational data, receive alarm information, and generate scheduling reports, enabling manual intervention and management of the system.

[0037] A mobile energy storage vehicle collaborative scheduling system based on NB-IoT also includes a security protection module. The security protection module adopts two-way authentication and data encryption technology to encrypt NB-IoT communication data and scheduling instructions to prevent data tampering and unauthorized access. At the same time, it sets up battery overcharge, over-discharge and over-temperature protection mechanisms. When abnormal parameters are detected, an alarm is automatically triggered and the charging and discharging circuit is cut off.

[0038] like Figure 3 As shown, the closed-loop collaborative scheduling workflow of the NB-IoT-based mobile energy storage vehicle collaborative scheduling system of this invention is as follows: Step 1, Data Acquisition: Each module of the perception layer collects the battery status and location information of the mobile energy storage vehicle, as well as the operating parameters and environmental parameters of the power distribution network nodes in real time, and encapsulates them into data frames through the NB-IoT terminal module; Step 2, Data Transmission: The NB-IoT terminal module uploads data to the communication gateway through the NB-IoT base station. After protocol conversion, the communication gateway forwards the data to the edge computing layer. Step 3, Edge Processing: The data preprocessing module of the edge computing layer performs noise reduction, normalization, outlier removal, and fusion processing on the data; the prediction error compensation module corrects the source-load prediction deviation; and the local collaborative decision-making module generates a short-term scheduling plan. Step 4, Global Scheduling: The edge computing layer uploads the processed data to the cloud scheduling layer. The multi-objective optimization scheduling module of the cloud scheduling layer constructs a multi-objective optimization model and uses an improved particle swarm optimization algorithm to solve the globally optimal scheduling scheme. Step 5, Instruction Issuance: The cloud scheduling layer 4 issues the global scheduling plan to the edge computing layer. The edge computing layer, in conjunction with the local scheduling plan, issues the scheduling instructions (charging and discharging power, driving route) to the mobile energy storage vehicle through the NB-IoT communication layer via the instruction issuance module. Step 6, Execution Feedback: The mobile energy storage vehicle executes the dispatch command and simultaneously uploads the execution status to the edge computing layer and the cloud dispatch layer through the NB-IoT communication layer; Step 7, Status Monitoring: The status monitoring module of the cloud scheduling layer monitors the system's operating status in real time. If an anomaly is detected, an alarm is triggered and staff are notified to handle it. Step 8, Rolling Optimization: Repeat the above steps every 15 minutes to update the data and scheduling scheme, and achieve closed-loop collaborative scheduling.

[0039] The mobile energy storage vehicle uses an NB-IoT-based mobile energy storage vehicle collaborative scheduling system to transport electricity from areas with surplus power (such as thermal power plants and wind power areas) to areas with high demand for power, and distribute the electricity to ensure a stable power supply for enterprises with power needs.

[0040] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT, characterized in that, It includes a perception layer, an NB-IoT communication layer, an edge computing layer, and a cloud scheduling layer, with each layer communicating with each other in sequence; The sensing layer includes a battery status monitoring module, a location positioning module, a power distribution network parameter acquisition module, and an environmental monitoring module. The battery status monitoring module is used to collect the voltage, temperature, state of charge, health status, and charge / discharge power parameters of the mobile energy storage vehicle's battery. The location positioning module is used to obtain the real-time location and driving trajectory of the mobile energy storage vehicle. The power distribution network parameter acquisition module is used to collect the voltage, power, and load demand data of the power distribution network nodes. The environmental monitoring module is used to collect the ambient temperature, humidity, precipitation, and meteorological parameters of the driving path and operating area.

2. The NB-IoT-based mobile energy storage vehicle collaborative scheduling system according to claim 1, characterized in that, The NB-IoT communication layer includes an NB-IoT terminal module, an NB-IoT base station, and a communication gateway. The NB-IoT terminal module is integrated into each acquisition unit of the perception layer and the mobile energy storage vehicle control terminal. It is used to encapsulate the acquired data into NB-IoT communication protocol data frames, upload the data to the communication gateway through the NB-IoT base station, and receive scheduling instructions issued by the edge computing layer and the cloud scheduling layer.

3. The mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 1, characterized in that, The edge computing layer includes a data preprocessing module, a local collaborative decision-making module, a prediction error compensation module, and an instruction issuance module. The data preprocessing module is used to perform noise reduction, normalization, and outlier removal on the data uploaded by the perception layer. The local collaborative decision-making module is used to realize short-term collaborative scheduling of multiple mobile energy storage vehicles within the region. The prediction error compensation module is used to correct the prediction deviations of photovoltaic / wind power output and load demand. The instruction issuance module is used to issue scheduling instructions to the mobile energy storage vehicles through the NB-IoT communication layer.

4. The NB-IoT-based mobile energy storage vehicle collaborative scheduling system according to claim 3, characterized in that, The local collaborative decision-making module of the edge computing layer uses a reinforcement learning algorithm to generate a short-term scheduling plan based on the real-time load of the distribution network and the status of mobile energy storage vehicles in the region. The prediction error compensation module employs scene generation and reduction technology, generating a typical source-load scene set through K-means clustering to correct prediction bias.

5. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 1, characterized in that, The cloud-based scheduling layer includes a data storage module, a multi-objective optimization scheduling module, a status monitoring module, and a human-machine interaction module. The data storage module stores historical data, real-time data, and scheduling schemes. The multi-objective optimization scheduling module constructs a multi-objective optimization model that includes minimizing voltage deviation, minimizing network loss, minimizing scheduling cost, and maximizing battery life, and uses an improved particle swarm optimization algorithm to solve for the optimal scheduling scheme. The status monitoring module monitors the operating status of the mobile energy storage vehicle and the distribution network in real time. The human-machine interaction module enables setting scheduling parameters, displaying scheduling schemes, and providing anomaly alarms.

6. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 2, characterized in that, It also includes a security protection module, which uses two-way authentication and data encryption technology to encrypt NB-IoT communication data and scheduling instructions to prevent data tampering and unauthorized access; at the same time, it sets up battery overcharge, over-discharge and over-temperature protection mechanisms, which automatically trigger an alarm and cut off the charging and discharging circuit when abnormal parameters are detected.

7. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 5, characterized in that, The cloud-based scheduling layer supports a rolling time-domain execution mechanism, which updates the source-load prediction data and mobile energy storage vehicle status data periodically, re-solves and optimizes the model, extracts and executes the scheduling instructions at the current moment, and forms a closed-loop feedback.

8. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 4, characterized in that, The local collaborative decision-making module specifically includes: A Markov decision process is constructed using a deep reinforcement learning algorithm, and its quintuple is defined as follows: in, The state space contains real-time state quantities such as the load of the regional distribution network, the output of distributed energy resources, the SOC and location of mobile energy storage vehicles, and the voltage of distribution network nodes. The action space includes scheduling actions such as adjusting the charging and discharging power and adjusting the driving path of the mobile energy storage vehicle; The state transition probability matrix is... Indicates the state Next action Then, the system transitions to state. The probability of; This is a discount factor with a value of 0.95, used to balance immediate rewards with long-term benefits; Here is a single-step reward function used to evaluate the quality of scheduling actions; the formula is: in, For node voltage deviation, For the increase in distribution network losses, For the total scheduling cost, These are weighting coefficients, with values ​​of 0.4, 0.3, and 0.3 respectively. The Q-learning algorithm is used for decision-making, and the temporal difference update formula for the Q value is: in, The learning rate is 0.

1. For state Next action Action value function; For the next state The maximum value of the action.

9. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 4, characterized in that, The prediction error compensation module specifically includes: Based on historical source load data, the prediction error of load and energy output is calculated: in, , They are respectively Load forecasting error and energy output forecasting error at any given time; , These are the actual values ​​for load and energy output, respectively. , These are the predicted values ​​for load and energy output, respectively, both in kW; Then, based on the probability distribution of the prediction error, a large number of source load error scenarios are generated. Subsequently, the K-means clustering algorithm is used to reduce the number of scenarios to obtain typical error scenarios. The objective function of clustering is: in, This is the clustering loss function; Number of typical scenarios; For the first A collection of scenarios for each cluster; The source-load error scenario sample vector; For the first The centroid vector of each cluster.

10. A mobile energy storage vehicle collaborative scheduling system based on NB-IoT according to claim 5, characterized in that, The objective function of the multi-objective optimization scheduling model used in the multi-objective optimization scheduling module is: in, For the scheduling period, , , , These are the weighting coefficients. For nodes exist Voltage amplitude at time 10:00 Rated voltage; For the line exist Current at any moment For the line The resistance, For the first The unit dispatch cost of a mobile energy storage vehicle For the first Mobile energy storage vehicles in The charging and discharging power at any given time For the first Mobile energy storage vehicles in The cost of battery degradation over time.

Images