CN-121983996-A - Photovoltaic reverse power absorption system based on mobile energy storage vehicle

CN121983996ACN 121983996 ACN121983996 ACN 121983996ACN-121983996-A

Abstract

The invention provides a photovoltaic reverse power consumption system based on a mobile energy storage vehicle, which is characterized in that a photovoltaic power generation unit is configured to output photovoltaic electric energy, supply power to loads and realize grid-connected operation through grid connection points, a reverse power detection and protection unit dynamically generates a self-adaptive reverse power protection threshold value through a time sequence prediction model, compares a real-time reverse power value with the reverse power protection threshold value, generates a reverse power trigger signal when the real-time reverse power value exceeds the threshold value and sends the reverse power trigger signal to a central control unit, the central control unit distributes a consumption task for the mobile energy storage vehicle cluster through a multi-agent collaborative decision model, plans and fuses a self energy state and a running path of running energy efficiency for each mobile energy storage vehicle participating in consumption, receives a scheduling instruction for each mobile energy storage vehicle in the mobile energy storage vehicle cluster, runs to a corresponding charging facility, and accesses a charging loop to consume the reverse power electric energy generated by the photovoltaic power generation unit.

Inventors

FENG XINYU
WU YING
MA XIN
Wang Dinggun
JI JIE
HUANG HUI
DING ZUJUN
LIU BAOLIAN
BAI QIUCHAN
ZHUANG XUZHOU

Assignees

淮阴工学院

Dates

Publication Date: 20260505
Application Date: 20260312

Claims (10)

1. The photovoltaic reverse power consumption system based on the movable energy storage vehicle is characterized in that a reverse power detection and protection unit is electrically connected with a photovoltaic power generation unit, a central control unit is respectively connected with the reverse power detection and protection unit, a charging facility and a movable energy storage vehicle cluster in a communication manner, and the charging facility is electrically connected with an electric energy output side of the photovoltaic power generation unit; The photovoltaic power generation unit is configured to output photovoltaic electric energy, power the load and realize grid-connected operation through grid connection points; The reverse power detection and protection unit is configured to collect historical operation data, a real-time output state, a load state and real-time environment parameters of the photovoltaic power generation unit, dynamically generate a self-adaptive reverse power protection threshold value through a time sequence prediction model, compare a real-time reverse power value with the reverse power protection threshold value, generate a reverse power trigger signal when the real-time reverse power value exceeds the threshold value, and send the reverse power trigger signal to the central control unit; the central control unit is configured to receive the reverse power trigger signal, allocate a digestion task to the movable energy storage vehicle clusters through the multi-agent collaborative decision-making model based on a real-time reverse power state, an available state of a charging facility and a real-time state of each movable energy storage vehicle, plan and fuse a running path of the self energy state and running energy efficiency for each movable energy storage vehicle participating in digestion, generate a corresponding scheduling instruction and send the scheduling instruction to the movable energy storage vehicle clusters; Each movable energy storage vehicle in the movable energy storage vehicle cluster is provided with an energy storage battery pack, a communication module, a positioning module and a driving unit, and is configured to receive the scheduling instruction, drive to a corresponding charging facility to access a charging loop, and consume reverse power electric energy generated by the photovoltaic power generation unit.
2. The mobile energy storage vehicle-based photovoltaic reverse power absorption system according to claim 1, wherein: The inverse power detection and protection unit comprises a data processing sub-module and a dynamic threshold prediction sub-module, wherein the data processing sub-module sequentially carries out moving average filtering and irradiance coupling correction preprocessing on collected original data, the moving average filtering is used for inhibiting high-frequency noise in the original data and outputting smooth photovoltaic power signals, the irradiance coupling correction is used for correcting photovoltaic power according to irradiance deviation and eliminating environmental interference, and the corrected photovoltaic output power meets the following expression: In the formula, The output power value of the photovoltaic after the moving average treatment at the moment t, For the correction coefficients obtained by regression fitting of the historical irradiance data, The real-time irradiation intensity at the time t, For the average value of irradiation intensities over the past preset 24 hours, For the standard deviation of irradiation intensity within 24 hours preset in the past, And (5) obtaining the photovoltaic output power value after irradiance coupling correction at the moment t.
3. The mobile energy storage vehicle-based photovoltaic reverse power absorption system according to claim 2, wherein: the time sequence prediction model adopted by the reverse power detection and protection unit is an LSTM model, the LSTM model builds a time sequence input vector based on the preprocessed corrected photovoltaic power, the preprocessed power change rate and the preprocessed real-time irradiation intensity, and outputs a self-adaptive reverse power protection threshold value, wherein the threshold value output meets the following expression: In the formula, For an adaptive inverse power protection threshold at time t, Is a hidden state vector at the moment of the LSTM model t, The weight parameters of the layers are output for the LSTM model, The bias parameters of the output layer are the LSTM model, In order to modify the linear cell activation function, As an adjustment factor for the rate of change of power, Is the real-time change rate of the output power of the photovoltaic.
4. A mobile energy storage vehicle-based photovoltaic reverse power absorption system according to claim 3, wherein: The reverse power detection and protection unit judges whether to generate a reverse power trigger signal through the following logic: In the formula, Indicating that the reverse power trigger signal is generated, Indicating that the inverse power trigger signal is not generated, The photovoltaic output power value after pretreatment and correction at the moment t, The real-time load power of the system at the time t, For the maximum reverse power value allowed by the grid, And outputting an adaptive inverse power protection threshold value for the LSTM model at the time t.
5. The mobile energy storage vehicle-based photovoltaic reverse power absorption system according to claim 1, wherein: The multi-agent collaborative decision model adopted by the central control unit is MADDPG algorithm, the MADDPG algorithm integrates photovoltaic point position parameters, charging facility state parameters and dynamic parameters of the movable energy storage vehicle into a global state space, and the global state vector meets the following expression: In the formula, Is the global state vector at time t, For the battery state of charge of all mobile energy storage vehicles, For the real-time reverse power value of each photovoltaic node, In order to be able to use the charging facility, For the coordinates of the obstacle within the dispatch area, Is the real-time output power of the photovoltaic power generation unit, For the charging power of the mobile energy storage vehicle, For the real-time travel speed of the mobile energy storage vehicle, To provide charging efficiency for the charging facility.
6. The mobile energy storage vehicle-based photovoltaic reverse power absorption system according to claim 5, wherein: the MADDPG algorithm constructs a multi-objective rewards function optimization scheduling policy that satisfies the following expression: In the formula, Is the prize value at time t, The sum of the photovoltaic reverse power actually consumed by all the movable energy storage vehicles at the current moment, The total generated power of the photovoltaic power generation unit at the current moment, For the battery state of charge of the ith mobile energy storage vehicle, For the movement distance of the ith movable energy storage vehicle at the current moment, As a weight coefficient of the charging efficiency, For the actual charging efficiency of the charging facility, And charging the movable energy storage vehicle in real time.
7. The photovoltaic reverse power consumption system based on the mobile energy storage vehicle according to claim 6, wherein the weight proportion of the photovoltaic consumption term, the charge state balance punishment term and the movement distance punishment term in the multi-objective rewarding function is determined through pareto front edge analysis.
8. The mobile energy storage vehicle-based photovoltaic reverse power absorption system according to claim 1, wherein: The central control unit plans a running path for fusing the self energy state and the running energy efficiency of the movable energy storage vehicle through a D-Lite algorithm, the algorithm sets a dynamic cost function for the running path, the battery charge state and the real-time running speed of the movable energy storage vehicle are fused into path cost calculation, and the dynamic cost function meets the following expression: In the formula, For the final dynamic cost from vertex v to vertex v', For the base move cost from vertex v to vertex v', As the state of charge weight coefficient, To be the real-time battery state of charge of the mobile energy storage vehicle, As a result of the velocity weighting factor, Is the real-time running speed of the movable energy storage vehicle.
9. The photovoltaic reverse power consumption system based on the mobile energy storage vehicle according to claim 8, wherein the D-Lite algorithm detects environmental changes in a scheduling area in real time, updates path costs and priorities of affected vertexes when path cost changes caused by obstacles are detected, corrects local traveling paths through incremental rescheduling, and triggers an exception handling mechanism when no feasible paths exist.
10. The photovoltaic reverse power consumption system based on the mobile energy storage vehicle, which is disclosed in claim 1, is characterized in that the charging facility collects the charging state of the mobile energy storage vehicle and the state of the energy storage battery pack in real time and feeds the charging state and the state of the energy storage battery pack back to the central control unit, and the running energy consumption of the running driving unit of the mobile energy storage vehicle is supplied by the energy storage battery pack carried by the mobile energy storage vehicle.

Description

Photovoltaic reverse power absorption system based on mobile energy storage vehicle Technical Field The invention belongs to the technical field of distributed photovoltaic power generation grid-connected operation control and energy storage system optimization scheduling, and particularly relates to a photovoltaic reverse power consumption system based on a mobile energy storage vehicle. Background With the continuous increase of the duty ratio of the photovoltaic power generation in an energy structure, the photovoltaic system is easy to trigger reverse power protection when in low load or power grid fault, so that a large number of light rejection problems are increasingly highlighted. The traditional solution mainly depends on a fixed energy storage system or an unloading resistor, has obvious limitations that the fixed energy storage device is generally deployed at a specific position in a container type or cabinet type form by adopting energy storage media such as lead-acid batteries and lithium ion batteries, the capacity and the power configuration of the fixed energy storage device are relatively fixed, the space-time fluctuation characteristics of distributed photovoltaic power generation are difficult to match, the unbalanced phenomenon that the local capacity is insufficient and the capacity of other areas is idle often occurs under the scene that the peak-valley difference of light Fu Chuli is large and the reverse power generation position is dispersed is difficult to occur, and the unloading resistor scheme is low in cost and quick in response, but converts precious electric energy into heat energy for dissipation, so that not only is the energy waste reduced, but also the overall economy of the system is reduced, the potential safety hazards such as heat dissipation and fire prevention are possibly caused, and the basic principle of high-efficiency utilization of green energy is not met. Particularly in the scenes of large-scale photovoltaic power stations, industrial parks, commercial building groups and the like, the photovoltaic array has wide distribution range and large load characteristic difference, and the instantaneous property, the intermittence and the space dispersibility of the reverse power have higher requirements on the application technology. In the prior art, a judgment mechanism based on a fixed threshold is mostly adopted for the inverse power detection, and a static power percentage or absolute value is usually set as a trigger threshold. The simple threshold method cannot adapt to rapid power fluctuation caused by abrupt change of illumination intensity and cloud cover, is difficult to respond to dynamic change of load, and is easy to generate two typical problems in actual operation, namely, when threshold setting is too conservative, reverse power can be actually exceeded by the power grid bearing capacity and is not timely triggered to be protected to influence stable operation of the power grid, and when threshold setting is too sensitive, frequent misoperation can be caused due to short fluctuation of illumination or instantaneous change of load, unnecessary system shutdown or frequent start-stop of energy storage equipment can be caused, equipment aging can be accelerated, and the effective photovoltaic power generation time can be reduced. In addition, the traditional detection method is mostly based on simple comparison of real-time power sampling values, lacks the capability of deep analysis of historical operation data and prediction of future trends, and is difficult to early warn and preprocess in the early stage of reverse power formation. In the aspect of energy storage scheduling, the existing scheme generally adopts a static strategy based on rules or simple priorities, for example, task allocation is carried out according to the residual capacity of the energy storage equipment, the distance or the fixed rotation sequence. The method lacks an overall optimization view angle for the collaborative operation of the multiple energy storage units, and cannot dynamically decide according to multidimensional information such as real-time inverse power distribution, states (such as charge states, health degrees and current positions) of all the energy storage devices, availability of charging facilities and the like. When multiple photovoltaic nodes simultaneously generate reverse power and the consumption demands compete, the static scheduling strategy often causes excessive use of part of energy storage equipment and idle of other equipment, or the uneconomic behavior that the energy storage equipment moves for a long distance to consume a small amount of reverse power occurs, and the maximization of the overall consumption efficiency of the system and the minimization of the operation cost are difficult to realize. Although there have been studies attempting to introduce optimization algorithms for task allocation, most focus on single vehicle or fixed path