Power grid demand response method and system based on dynamic optimization of autonomous driving bus driving strategy

Abstract

The present application relates to the technical field of smart grid demand response, in particular to a power grid demand response method and system based on dynamic optimization of automatic driving bus driving strategy, vehicle state is collected to determine the station adjustment interval and screen the response trajectory, combined with load monitoring and demand instruction to determine the power grid load limiting period, according to the relationship between the predicted station and the limiting period, the trajectory is classified and labeled as access type, and then the vehicle driving and parking rhythm is adjusted to generate charging time sequence sorting, combined with real-time load verification, the actual access sequence of the vehicle is generated and the execution record is generated. The present application generates the station adjustment range based on the driving state and the electric quantity, arranges the charging sequence combined with the power grid load fluctuation, charges in advance or delays in the limited period, sorts the charging behavior and identifies the abnormality, realizes the accurate matching of charging and load constraint, reduces the peak pressure and improves the electric energy efficiency, guarantees the task completion and reduces the conflict, forms the adaptive mechanism, realizes the dynamic matching optimization of charging and power grid.

Description

Technical Field

[0001] This invention relates to the field of smart grid demand response technology, and in particular to a grid demand response method and system based on dynamic optimization of autonomous driving bus driving strategies. Background Technology

[0002] The field of smart grid demand response technology includes power system load regulation methods, electricity consumption-side resource coordination mechanisms, grid operation status monitoring, and distributed energy access management. The core of this technology lies in identifying the load characteristics of electricity users and combining them with grid operation constraints to achieve dynamic matching of power supply and demand. Its overall technical system covers grid load forecasting, data acquisition and communication mechanisms, price incentive mechanisms, electricity consumption behavior response models, and collaborative control strategies for multiple types of energy-consuming equipment. By modeling the correlation between the operating status of user-side equipment and grid-side dispatching demand, it achieves time-sharing regulation and optimized allocation of power load.

[0003] Among them, the grid demand response method and system based on dynamic optimization of autonomous bus driving strategy refers to a technical solution that uses the driving path planning, stop time arrangement and charging behavior control of autonomous buses during operation to associate with grid load dispatching demand. It involves real-time location data acquisition of autonomous buses, driving speed sequence recording, stop time sequence generation and vehicle battery remaining power detection. By constructing a mapping relationship between bus operation timetable and grid load curve, and combining charging time window division, charging power allocation rules and vehicle dispatching order adjustment methods, the charging behavior of buses in different time periods is constrained and guided, thereby forming a grid load response control process based on bus operation strategy.

[0004] Existing technologies cannot dynamically identify vehicle driving stability and flexibly adjust charging time in actual operation. They lack the ability to judge in real time whether the power supply meets the target charging demand. There is a risk of conflict between charging sequence and grid load limitations. They cannot flexibly adjust charging access for periods of load limitation, resulting in difficulty in reducing peak load or some vehicles waiting to charge. Charging scheduling lacks a matching strategy that is adaptive to driving status, time window and power supply conditions. During the execution process, charging conflicts or delays are likely to occur, reducing the accuracy of overall demand response and the effect of grid load regulation. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide a power grid demand response method and system based on dynamic optimization of autonomous driving bus driving strategies. The technical solution is as follows: A power grid demand response method based on dynamic optimization of autonomous bus driving strategies includes the following steps: S1: Collect the vehicle speed sequence, real-time location, predicted arrival time and battery status of autonomous buses, read the planned arrival time, identify stable driving sections and trajectory sections, mark the time offset status, determine the arrival time adjustment interval, verify the battery availability status, and filter to obtain the set of trajectories that can participate in demand response. S2: Determine the distribution node corresponding to the grid access location of the target charging station, obtain the load change sequence and sampling time, read the demand response command, identify the continuous load-limited section according to the sampling time, compare it with the load-limited time period, merge the continuous overlapping time periods, and obtain the grid load-limited time period. S3: Extract the predicted arrival time and arrival time adjustment interval from the set of available demand response trajectories, read the start and end times of the power grid load restriction time segment, mark the three access types of early, late and sequential according to the time position relationship, and sort them to obtain the charging access timing adjustment set. S4: Based on the access type flag of the charging access timing adjustment set, determine the vehicle driving and stopping adjustment mode, and control the vehicle driving forward, backward or maintain the original rhythm through vehicle control commands. According to the adjusted arrival time, access time and stopping order, obtain the bus charging time sorting result. S5: Based on the bus charging time sorting results, read the real-time load change sequence of the determined power distribution nodes, mark the real-time load-limited sections, verify the vehicle charging access time and actual access sequence, record compliant charging behavior, summarize abnormal charging behavior, and obtain the power grid demand response execution record.

[0006] A power grid demand response system based on dynamic optimization of autonomous driving bus driving strategies, the system comprising: The trajectory filtering module collects the vehicle speed sequence, real-time location, predicted arrival time and battery status of autonomous buses, reads the planned arrival time, identifies stable driving sections and trajectory sections, marks the time offset status, determines the arrival time adjustment interval, verifies the battery availability status, and filters out a set of trajectories that can participate in demand response. The load identification module determines the distribution node corresponding to the grid connection location of the target charging station, obtains the load change sequence and sampling time, reads the demand response command, identifies continuous load-limited sections according to the sampling time, compares them with the overlapping load-limited time periods, merges the continuous overlapping time periods, and obtains the grid load-limited time period. The access classification module extracts the predicted arrival time and arrival time adjustment interval from the set of available demand response trajectories, reads the start and end times of the power grid load restriction time segment, marks the three access types as early, late, and sequential according to the time position relationship, and sorts them to obtain the charging access timing adjustment set. The rhythm control module determines the vehicle driving and stopping adjustment mode based on the access type mark of the charging access timing adjustment set. Through vehicle control commands, it controls the vehicle to move forward, delay, or maintain the original rhythm. Based on the adjusted arrival time, access time, and stopping order, it obtains the bus charging time sorting result. The execution verification module reads the real-time load change sequence of the determined power distribution nodes based on the bus charging time sorting results, marks the real-time load-limited sections, verifies the vehicle charging access time and actual access sequence, records compliant charging behavior, summarizes abnormal charging behavior, and obtains the power grid demand response execution record.

[0007] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: This invention can identify stable driving sections and generate arrival time adjustment ranges based on real-time vehicle driving status and power limitations. It can also rationally arrange the charging access sequence in combination with grid load fluctuations, enabling early or delayed charging during periods of load constraint. By organizing charging behavior according to time sequence and identifying abnormal access, it achieves precise matching between charging behavior and load constraints, reduces peak load pressure and improves energy utilization efficiency, ensures that vehicles complete charging tasks as planned and reduces waiting conflicts, enhances the synergy between charging scheduling and grid load regulation, and forms an adaptive charging control and sequencing mechanism based on driving status, power level and load conditions, thereby achieving dynamic matching and optimization between the charging process and grid operation. Attached Figure Description

[0008] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the acquisition process of S1 in this invention; Figure 3 This is a flowchart illustrating the acquisition process of S2 in this invention; Figure 4 This is a flowchart illustrating the acquisition process of S3 in this invention; Figure 5 This is a flowchart illustrating the acquisition process of S4 in this invention; Figure 6 This is a flowchart of the acquisition process for S5 of the present invention. Detailed Implementation

[0009] The following detailed explanation illustrates the specific implementation methods: The basic implementation examples are as follows: Figure 1 As shown: A power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy includes the following steps: S1: Collect the speed change sequence, real-time vehicle location, predicted arrival time, and vehicle battery status of the autonomous bus; read the planned arrival time from the bus operation plan; organize the speed change sequence according to time sequence; determine the stable driving segment when the speed fluctuation is within a preset range and lasts for a preset duration; determine the corresponding trajectory segment by combining the real-time vehicle location; compare the predicted arrival time with the planned arrival time; mark the time offset status according to the relationship of early, on-time, and late; determine the earliest and latest arrival times that the predicted arrival time can be adjusted based on the time offset status and the duration of the stable driving segment; form the arrival time adjustment interval; determine whether the remaining power meets the minimum power requirement to reach the target charging station based on the vehicle battery status; mark the power availability status; filter the trajectory segments that have the arrival time adjustment interval and meet the power availability status requirements; and obtain the set of trajectories that can participate in demand response. The set of trajectories that can participate in demand response includes trajectory segments, allowable adjustment range of arrival time, battery availability status, and time offset status.

[0010] Please see Figure 2 Specifically, S1 includes: S11: Collect the speed change sequence, real-time vehicle location, predicted arrival time, and vehicle battery status of the autonomous bus; read the planned arrival time from the bus operation plan; organize the speed change sequence according to the operation sequence; verify whether the speed fluctuation is within the preset range; delineate the driving trajectory segment by associating the real-time vehicle location; and obtain the stable trajectory segment. Details are as follows: The system collects data on the speed change sequence, real-time vehicle location, predicted arrival time, vehicle battery status, and planned arrival time from the bus operation plan for autonomous buses. It also reads high-frequency pulse signals from Hall effect wheel speed sensors via the onboard CAN bus and converts them into a speed change sequence by the vehicle controller. The vehicle's real-time location coordinates are obtained through an onboard GNSS / BDS dual-mode high-precision positioning module. The predicted arrival time is retrieved from the dispatch center server via the OTA interface of the vehicle-mounted 4G / 5G communication terminal based on the TCP / IP communication protocol. With the planned arrival time The vehicle battery status is collected through the internal register interface of the BMS battery management system. The acquired vehicle speed change sequence was processed according to... The sampling periods are arranged numerically, and the maximum vehicle speed within the current time window is retrieved. and minimum value The fluctuation difference is obtained by performing a subtraction operation. The difference With preset fluctuation threshold Perform size comparison, where Set as This setting is based on the standard cruise speed difference envelope of urban buses on non-congested dedicated lanes, and simultaneously retrieves statistics from internal timers. Number of consecutive sampling points that meet the conditions Calculate duration ,like Reaching the preset time threshold Then, this sampling sequence is extracted as a stable observation window. For example, in a certain BRT lane operation, the vehicle speed is... From the inside Slow fluctuations to Its fluctuation difference is Less than the set value If the system meets the stability criteria, and the fluctuation difference exceeds the preset fluctuation threshold (indicating inconsistent continuity), it indicates that the vehicle is in a state of frequent starts and stops or congestion. The system directly marks the current time window as unstable, clears the internal timer, and slides the step size to the next sampling period for recalculation, based on the starting sampling sequence number corresponding to the stable observation window. and end sampling number From the real-time location set of vehicles The coordinates of the indexed point, such as from the sequence number coordinates of Extract to serial number coordinates of This continuous set of latitude and longitude points in physical space is defined as the corresponding trajectory segment, and the predicted arrival time is retrieved. With the background planning time Perform timestamp alignment and association, complete data initialization and loading, and encapsulate the extracted data packets containing time-series vehicle speed, physical trajectory, battery information and arrival time deviation into a structured dictionary to obtain stable trajectory segments.

[0011] S12: Retrieve the stable trajectory segment, predicted arrival time, and planned arrival time; mark the arrival time as ahead, on time, or delayed according to the order of arrival; and, considering the continuity of the stable travel segment, determine the allowable forward and backward movement boundaries of the predicted arrival time, generating the arrival time adjustment interval; details are as follows: Call the predicted arrival time corresponding to the stable trajectory segment Planned arrival time and the duration of stable driving process Execution time difference calculation formula ,like Negative value and absolute value greater than If so, the trajectory is marked with the value "1" in the status register to represent the advanced state. Positive value and greater than If the absolute value is within 3, it represents a delayed state. If the value is "2", it indicates the on-time status, for example, the predicted arrival time is... The planned destination is The difference is obtained by subtraction. The condition is determined to be in an advanced state based on the duration of the stable driving section. Calculate the allowable adjustment boundary values; Set forward coefficient and shift coefficient The specific settings are based on the upper limit of the redundancy of the electronic traction control adjustment and the optimal range of the braking energy recovery efficiency of the autonomous driving system on a specific road segment, and the boundary is shifted forward through calculation. Assuming the above-mentioned data collection Multiplication and subtraction operations allow shifting to the left. The calculated forward boundary is Similarly, the boundary shift can be calculated using formulas. Allow to move backward The calculated shifted boundary is as follows: If the calculated To ensure on-time arrival, the original predicted time remains unchanged, and the adjustment interval is not expanded. The obtained forward boundary time is taken as the earliest time that the predicted arrival time can be adjusted, and the backward boundary time is taken as the latest time. The two together form a closed interval. During this process, each stable trajectory segment within the bus is traversed and calculated, and the clock offset is superimposed for each time node. The obtained time boundary values ​​are written into the time axis mapping database in the form of key-value pairs to generate the arrival time adjustment interval.

[0012] S13: Based on the arrival time adjustment interval and vehicle battery status, read the minimum battery requirement of the target charging station, verify the charging station access requirements corresponding to the remaining battery, mark the battery availability status, and filter the trajectory segments with arrival time adjustment intervals and battery availability status that meet the requirements to obtain the set of trajectories that can participate in the demand response; as detailed below. Based on the arrival time adjustment range and the vehicle's current battery status and the remaining driving distance of the vehicle to the target charging station. Retrieve the minimum safe power threshold set for the target charging station from the charging station management platform. The threshold is set to The settings are based on the inflection point of the voltage drop curve of lithium iron phosphate power batteries in the low-charge range and the physical protection limits to avoid over-discharge. The underlying computing resources are used to calculate the expected charge level when arriving at the charging station using energy consumption prediction logic. ; Its calculation logic is as follows , in the formula The percentage of battery power remaining after reaching the target station. This represents the current battery percentage collected by the vehicle's BMS. The unit of distance for the in-vehicle navigation module to plan the path to the charging station based on the GIS map is... , The average energy consumption per unit mileage for this vehicle model is calculated by accessing historical vehicle power consumption data. , The total capacity on the battery nameplate is in units of ,multiplier Used to convert decimals to percentage proportions, such as the current for ,distance for , for , for Execute the calculation process and This yields the power consumption ratio, allowing for further execution. To obtain the expected power consumption, the calculated With safety threshold Perform size comparison operations, if If the battery status is deemed satisfactory, the available battery status bit in the feature vector is set to "1". If the expected battery level is less than [a certain value], the battery status is determined to be satisfactory. If the situation indicates a risk of mid-journey breakdown, the bit is set to "0". Then, a multi-criteria parallel filtering logic is executed, retrieving all generated trajectory data packets and checking if they simultaneously possess a non-empty time adjustment interval subset and a "1" flag indicating available battery power. Packets with a "0" flag, meaning the predicted battery power is lower than [a certain value], are removed. For unstable road segments where frequent congestion makes it impossible to calculate adjustment intervals, the remaining trajectory index numbers that meet the dual conditions are grouped together. By sorting the arrival times of each grouped trajectory segment in ascending order, a dynamically updated resource pool with actual scheduling redundancy is extracted, resulting in a set of trajectories that can participate in demand response.

[0013] S2: Determine the corresponding distribution node based on the grid access location of the target charging station, obtain the load change sequence and corresponding sampling time output by the load monitoring equipment of the distribution node according to the preset sampling period, read the load restriction period and restriction threshold in the demand response command issued by the grid side, organize the load change sequence according to the sampling time, determine the time period when the load value corresponding to adjacent sampling time reaches the restriction threshold or above as the continuous load-limited section, compare the continuous load-limited section with the load-limited period, retain the overlapping time period, merge the overlapping time periods with continuous time and interval not exceeding the preset sampling period, and obtain the grid load-limited time section; The power grid load restriction time period includes continuous load restriction period, load restriction overlapping period, and merged continuous period.

[0014] Please see Figure 3 Specifically, S2 includes: S21: Determine the corresponding distribution node based on the grid connection location of the target charging station, obtain the load change sequence and corresponding sampling time output by the load monitoring equipment of the distribution node according to the preset sampling period, read the load restriction period and load restriction threshold in the demand response command issued by the grid side, organize the load values ​​according to the sampling time, compare the sampling time boundary with the load restriction period boundary, and obtain the node load reference time sequence; as detailed below: Based on the geographical coordinates of the target charging station Retrieve the power grid distribution GIS topology table, locate the distribution node number "DN-Bus-Charging-01" whose power supply radius covers the coordinates, and use the fiber optic communication interface based on the IEC61850 standard communication protocol to access the real-time database of the front-end smart meter AMI and SCADA system of this distribution node to extract its historical active power load sequence. ; Synchronously extract the timestamp sequence corresponding to each power value. The sampling period Fixed as The demand response command message is read from the data interface of the superior power grid dispatch center, and the start time of load limiting is obtained by parsing it. End time and load limit threshold Retrieve the rated capacity from the nameplate of the transformer at this distribution node. and the current power factor as monitored Perform multiplication operations The rated active power of the node is obtained as follows Call the preset safety operation coefficient This coefficient is set with reference to the transformer oil temperature during the summer peak period. to The maximum safe load rate at any time; Through formula The calculated threshold is The acquired load sequences are clock-aligned according to the sampling time. For each set of sampled data, its sampling time is retrieved. respectively with and Perform a timestamp size comparison, if and Then mark the sampling point as the evaluation target, for example... Corresponding load value Perform subtraction operation The difference obtained is greater than The sampling points and their associated time coordinates are stored in a structured manner. If the difference is less than or equal to... This indicates that the limit has not been exceeded, and it is marked as safe and not recorded. This is achieved through iteration. All sampling points within an hour are selected, and all load fluctuation sample sequences that fall within the demand response period and exceed the limit are identified to obtain the node load reference time series.

[0015] S22: The load change sequence, corresponding sampling time, and load limit threshold in the node load reference time series are called. Paired load values ​​are read according to adjacent sampling times. Each pair of load values ​​is compared with the load limit threshold. The time period in which both load values ​​reach or exceed the load limit threshold is selected and compared with the load limit time period to obtain the load limit overlap segment; specifically as follows: The call node load reference time series includes the load change sequence, sampling timestamps, and calculated load limit thresholds. Set the loop pointer Iterate starting from the first sampling sequence number within the demand response period, reading pairs of adjacent load values. For example, reading load value and load value ,Will and If the comparison operation determines that the value is greater than the threshold, then... and The comparison operation determines if a value is greater than a threshold; if both adjacent values ​​are strictly greater than a threshold... When the condition of continuous exceeding the limit is met, a time period extraction operation is performed, and the start time of the segment is recorded as [time value]. The end time is If it appears and ,because If the continuity is broken, the sampling interval is not included in the restricted segment. Boundary values ​​are extracted for all selected time periods that meet the conditions, and denoted as... Invoke the global time limit in the demand response instruction. Perform interval intersection operation, specifically by calling the maximum value function. Determine the time value of the overlap start point and call the smaller function. Determine the time value of the overlapping endpoint, such as for a segment. The calculation yields the overlap start point as The overlapping endpoint is By performing the above intersection verification on all sample pairs that meet the double-value over-limit conditions, invalid intervals with a starting point greater than or equal to the ending point after the intersection operation are removed, i.e., scattered time slices that are not within the load-limited period. All valid time intervals retained by the intersection operation are stored in a dynamic array in memory to obtain the load-limited overlapping segment.

[0016] S23: Based on the load-limiting overlapping segment, the preset sampling period, and the corresponding sampling time, arrange the overlapping time periods in chronological order of their start times. Compare the interval between the end time of adjacent overlapping time periods and the next start time. Merge overlapping time periods that are continuous in time and whose interval does not exceed the preset sampling period. Record the start and end times of the merged period to obtain the power grid load-limiting time segment; specifically as follows: Based on load-limited overlapping sections and preset sampling period The starting and ending sampling times for each segment are also considered. All overlapping time segments in the dynamic array are processed using a bubble sort algorithm from earliest to latest starting time. Two adjacent time segments after sorting are extracted, and the first segment is defined as... The second paragraph is Retrieve the end time of the first segment With the start time of the second paragraph ; Calculate the time interval for performing subtraction. The difference obtained With preset sampling period Perform size determination, because If the two segments are determined to have a continuous temporal influence, a segment merging operation is performed, and the start time of the first segment is merged. As the starting point of a new time period, the end time of the second segment will be... As the end of the new time period, update the current merged segment to... It also releases the original two memory nodes and reads the third time segment from the array. The subtraction operation is called to calculate the interval between the end of the current merged segment and the start of the third segment. ,Will and Comparison, because If the two segments are determined not to belong to the same continuous restricted process, and forced merging would lead to misjudgment of excessively long unrestricted periods, then the merging operation of the current element is stopped, and the segments are solidified into independent intervals. A new merging benchmark is started. By performing the above recursive difference comparison and boundary merging operation on all elements in the set, the fragmentation problem of restricted segments caused by instantaneous load sampling fluctuations in the distribution network is eliminated. The start and end boundary times of each independent continuous time period are finally determined and written into the distribution network operation constraint data table in a unified format to obtain the power grid load restriction time segment.

[0017] S3: Extract the predicted arrival time and arrival time adjustment interval corresponding to each trajectory segment in the set of trajectory segments that can participate in the demand response. Read the start and end times of each grid load restriction time segment. Use the time position relationship between the predicted arrival time and the start and end times as the classification basis to mark the access type of each trajectory segment. The trajectory segment with the predicted arrival time not earlier than the start time and not later than the end time is marked as delayed access type. The trajectory segment with the predicted arrival time earlier than the start time and the end time of the arrival time adjustment interval not later than the start time is marked as early access type. The trajectory segment with the predicted arrival time later than the end time is marked as sequential access type. Organize the trajectory segments of each type according to the order of the predicted arrival time to obtain the charging access timing adjustment set. The charging access timing adjustment set includes early access trajectory, delayed access trajectory, sequential access trajectory, and trajectory arranged according to the predicted arrival order.

[0018] Please see Figure 4 Specifically, S3 includes: S31: Extract the predicted arrival time and arrival time adjustment interval for each trajectory segment in the set of trajectory segments eligible for demand response. Read the start and end times of each power grid load limitation time segment. Associate the predicted arrival time, adjustment interval boundary, and load limitation boundary by trajectory segment number. Compare the sequential relationship between the predicted arrival time and the load limitation boundary to obtain the access location relationship value; specifically as follows: Extract the predicted arrival time and arrival time adjustment range for each trajectory segment in the set of trajectory segments eligible for demand response, and retrieve the trajectory number from memory address using the index pointer. Associated predicted arrival time values Simultaneously read the lower limit of its arrival time adjustment range. and upper limit retrieve track number Predicted time and its interval retrieve track number Predicted time and its interval Retrieve the load limit start time of node number "DN-Bus-Charging-01" from the distribution network operation constraint data table. and end time This converts all time strings into integer seconds, calculated from zero, using a conversion operator. Calculated as , Calculated as Get the conversion result , , and limiting boundaries , Perform a time association operation on each trajectory number, and... Data rows and and Perform horizontal splicing comparison, targeting Calling the subtraction operator to execute Execute again ,against implement ,against implement The logical judgment operator is invoked to evaluate the signs of the two sets of differences mentioned above. If the difference between the predicted time and the starting time is... Or a positive number and a negative difference from the end time or If the difference between the predicted time and the start time is negative, the value "-1" is written to the access location relationship value register to indicate that it is inside the restricted interval. If the difference between the predicted time and the end time is positive, the value "1" is written to indicate that it is later than the restricted interval. The access location relationship value is obtained by performing a cyclical subtraction comparison on all trajectories in the resource pool.

[0019] S32: Call the access location relationship value, predicted arrival time, end time of the arrival time adjustment interval, start time of load limitation, and end time of load limitation; determine three relationships: the predicted arrival time is within the load limitation interval, earlier than the start time of load limitation, or later than the end time of load limitation; and mark these relationships as delayed access type, early access type, and sequential access type, respectively, generating an access type label; details are as follows: The system retrieves the access location relationship value, predicted arrival time, end time of the arrival time adjustment interval, start time of load limit, and end time of load limit, for the trajectory where the location relationship value in the register is "0". retrieve its predicted time With load limiting range Perform boundary inclusion verification to determine if it satisfies the Boolean logic that it is no earlier than the start time and no later than the end time. Write the string code "delayed access" to the type identifier bit to wait for the restriction period to end. This applies to trajectories with a position relationship value of "-1". Retrieve the end time of its adjustment interval, i.e., the upper limit. With load limit start time Perform a timestamp subtraction comparison operation. Earlier And the difference interval is If the adjusted latest boundary is still strictly no later than the load limit start point and there is no overlap, write the string code "early access" in the type identifier bit. If the difference interval is less than If an overlap occurs, the trajectory will be downgraded; this applies to trajectories with a positional relationship value of "1". retrieve its predicted time With the end time of load limit Perform a numerical value determination; because To completely bypass the restriction period, the string code "sequential access" is written in the type identifier. To quantify classification accuracy and avoid time fragmentation errors, a time deviation coefficient is set. This coefficient is set with reference to the minimum time granularity allowed for scheduling by the electronic bus stop sign. Seconds, combined with this coefficient, are used to verify the mutual exclusion and uniqueness of each classification label through Boolean arithmetic, ensuring... Uniquely classified to the delayed group, Classified to the advance group, The data is classified into a sequential group. The determined classification result string is bound to the original trajectory unique number through a hash operation. A new feature column is created in the memory database for persistent storage, and an access type tag is generated.

[0020] S33: Based on the access type marker and access location relationship value, read the corresponding trajectory segments of delayed access type, early access type and sequential access type, organize the trajectory segments of each type in the order of predicted arrival time, check the correspondence between trajectory segment number, arrival time adjustment interval and power grid load restriction time segment, and obtain the charging access timing adjustment set; The trajectory segments of each type are organized in chronological order of predicted arrival times as follows: According to the access type marker, read the trajectory segments corresponding to delayed access type, early access type and sequential access type respectively, and maintain the correlation between the predicted arrival time, arrival time adjustment interval and power grid load restriction time interval corresponding to the trajectory segment, and form the intra-type arrangement record according to the order of the predicted arrival time; The specific steps for verifying the correspondence between the trajectory segment number, the arrival time adjustment interval, and the power grid load restriction time segment are as follows: The trajectory segment numbers are matched one by one with the arrival time adjustment interval, the power grid load restriction time interval, and the access location relationship value. The trajectory segments whose access type marker matches the access location relationship value are retained, and the retained trajectory segments are written into the charging access timing adjustment set; as follows: Based on the access type flag and access location relationship value, the trajectory that has been successfully marked as "early access" is read. Trajectories marked as "delayed access" And the trajectory marked as "sequential access" Retrieve the original predicted arrival times for each trajectory. In the scheduling system, a temporary doubly sorted linked list is established to... , as well as The system uses a bubble sort algorithm to process data according to the ascending order of timestamps. If a conflict occurs with the same timestamp, the remaining SOC (State of Charge) value of the corresponding vehicle is used as the secondary key for ascending arbitration to confirm the final sorting. Ranked first In the middle of the order. When it is at the end, execute the secondary verification logic for the time slack, and call... Corresponding arrival time adjustment range The intersection verification command is invoked to verify the grid load limiting time segment associated with this interval. ,confirm Although the predicted points overlap, their physical charging behavior can be delayed through parking and holding strategies, and the following can be retrieved. corresponding upper limit of the interval Its load limit start time By determining the magnitude of the bias, it is confirmed that even if it is at the latest point in the interval, it can completely avoid the restricted time period. By performing pointer sealing operations on each sorted trajectory node, the unique VIN identification code of each vehicle, the currently determined access type, the start and end boundary values ​​of the allowed adjustment time window, and the load constraint boundary of the target charging node are encapsulated into a time sequence scheduling control word with a unified format. The encapsulated data objects are then pushed into the edge computing execution buffer of the charging station management system in batches to complete the rearrangement and data locking of buses with different response capabilities in the time dimension, and obtain the charging access time sequence adjustment set.

[0021] S4: Based on the access type markers of each trajectory segment in the charging access timing adjustment set, determine the vehicle driving and stopping adjustment methods, and generate control commands through the on-board driving controller. For the early access type, control the vehicle driving rhythm to move forward, and ensure that the adjusted predicted arrival time is within the arrival time adjustment interval and not later than the start time of the corresponding grid load restriction time segment. For the delayed access type, control the vehicle stopping rhythm to be delayed, and ensure that the adjusted charging access time is not earlier than the end time of the corresponding grid load restriction time segment. For the sequential access type, maintain the original driving and stopping rhythm, and organize the vehicle charging sequence according to the adjusted predicted arrival time, charging access time, and stopping order to obtain the bus charging time sorting result. The bus charging time ranking results include the adjusted predicted arrival time, charging access time, vehicle charging sequence, and stop order.

[0022] Please see Figure 5 Specifically, S4 includes: S41: Based on the access type markers of each trajectory segment in the charging access timing adjustment set, read the trajectory segment number, predicted arrival time, arrival time adjustment interval, and grid load limitation time interval. Determine whether to advance the driving rhythm, delay the stopping rhythm, or maintain the original rhythm according to the access type markers. Call the on-board driving controller to write the corresponding control content and generate vehicle control commands; specifically as follows: Based on the charging access timing, adjust the access type label of each trajectory segment in the set, and extract the trajectory number from the edge computing execution buffer. Predicted arrival time Arrival time adjustment range Power grid load restriction time period And its corresponding "delayed access" marker, synchronously extracting the trajectory number. Predicted arrival time Arrival time adjustment range and its "early access" flag, for those marked as "early access" The remaining trajectory segment length can be retrieved via the vehicle-mounted T-Box. And the current vehicle speed fed back by the wheel speed sensors The command to advance the driving rhythm is executed, generating an action that calls a subtraction operation to determine the target arrival time that must be reached ahead of schedule. The formula is called to calculate the required average vehicle speed. ,in Let's assume the absolute starting time of the vehicle entering this stable trajectory is recorded as the time difference. ; Perform operations Right now , calculate The speed setpoint is encapsulated into a message data frame based on the CAN 2.0B standard protocol and written to the speed cruise control register of the vehicle control unit (VCU). This applies to settings marked as "delayed access". The command to delay the docking rhythm is executed, and the waiting time is accumulated by adding an additive operation to the original scheduled start time of refueling after arrival at the station. The calculated allowable time for the physical charging gun to lock and unlock is... The delay parameter command is sent to the parking charging and distribution system management module of the vehicle-mounted VCU for the trajectory marked as "sequential access". The system generates instructions to maintain the original cruise power and regular parking settings. It retrieves the corresponding binary action control word from the system dictionary, such as "move forward" corresponding to "01", "delay" corresponding to "10", and "keep" corresponding to "00". The control word is concatenated and encapsulated with parameters such as the trajectory number to form a control instruction data packet with a cyclic redundancy check (CRC) code, thus obtaining the vehicle control instructions.

[0023] S42: Invoke vehicle control commands, arrival time adjustment intervals, and the start and end times of the grid load restriction time segment. For early access types, adjust the predicted arrival time to be within the arrival time adjustment interval and no later than the start time of the corresponding grid load restriction time segment. For delayed access types, adjust the charging access time to be no earlier than the end time of the corresponding grid load restriction time segment. For sequential access types, maintain the original driving and stopping rhythm to obtain the access time boundary interval; details are as follows: Invoking vehicle control commands, arrival time adjustment intervals, and the start time of power grid load restriction time segments. and end time For trajectory number For the "early access" type, extract its original predicted arrival time. and adjustment range Retrieve sampling accuracy coefficients from configuration items Perform a step-by-step sliding window search within the adjustment range, Perform a forward subtraction offset operation using the reference point, setting a fixed offset. The adjusted target arrival time was calculated to be Call the comparison operator to verify Is it within the range? Within this timeframe, it is also verified whether the value is less than the load limit start time. After logical comparison, both conditions are met, and the time is determined to be... The precise charging access time, based on the trajectory number. For the "delayed access" type, retrieve its original predicted arrival time. With the end time of load limit ; Perform maximum value extraction operation The earliest lower limit time for obtaining permission to access is To mitigate the edge impact effect of sudden increases in grid load, a safety buffer duration parameter is introduced. This parameter is set based on the historical peak load decline rate of the distribution node. Perform addition operation The judgment operator is called to determine whether the value is within the maximum delay range allowed by the charging station's operation schedule. Marked as The precise access time, for Directly read and use its original predicted time. As an access point, the calculated and verified precise time values ​​are time-aligned in memory according to vehicle number, and a front and rear width of [value missing] is set for each time point. The allowable time fluctuation is used to construct the time window for each vehicle's occupancy operation at the energy replenishment interface, thus obtaining the access time boundary interval.

[0024] S43: Based on the access time boundary interval, vehicle control commands, and trajectory segments of sequential access types, read the adjusted predicted arrival time, charging access time, and stop order relationship. Arrange the charging order of vehicles with early access type, late access type, and sequential access type according to the time sequence to obtain the bus charging time sorting result; as follows: Based on the access time boundary interval, vehicle control commands, and sequential access type trajectory segments, the adjusted data is read sequentially from the memory stack in the previously ordered manner. Predicted access time Adjusted Charging access time And the original prediction time is of The trajectory information is retrieved by extracting the stop order index numbers of each vehicle as set in the initial bus operation plan. A quicksort algorithm is then used to sort the extracted access times in ascending order using only pure numbers, transforming them into... Right now corresponding The first execution priority assigned to the charging queue will Right now corresponding Assigned to second priority Right now corresponding Assigned to the third priority, during the priority arrangement and confirmation process, the idle status bit matrix of the physical charging piles in the charging station is retrieved through the interface. The overlap between the planned access time of each vehicle and the occupancy time of the preceding vehicle at the same pile position is verified, and a correction calculation formula including dimensional consistency is applied: ,in and The planned connection time for the two vehicles in front and behind. and To calculate the estimated end time of charging based on the battery demand of each vehicle, if the calculation result... If a physical time overlap conflict exists, then the access time of subsequent vehicles will be delayed by a cyclic addition step-by-step process until... For example, given the preorder inequality exist Connect and requires charging Then its expected end time for The following sequence plan Access and expected End, calculate the overlap between the two: ; because If there is no overlap and no physical conflict, the current ranking result is maintained. The final determined vehicle number sequence, the precise access timestamp, and the charging power limit value allocated according to the node capacity are concatenated into a data linked list to form a complete machine-readable charging station operation scheduling schedule. This schedule is then distributed to the charging station management system backend and the human-machine interaction terminal of each vehicle to obtain the bus charging time ranking result.

[0025] S5: Based on the vehicle charging sequence in the bus charging time sorting results, read the real-time load change sequence generated by the determined distribution nodes during the charging execution phase, and organize the real-time load change sequence according to the time sequence. According to the limit threshold in the demand response instruction, the time period when the real-time load value reaches the limit threshold is marked as the real-time load restricted section. The time position relationship between the charging access time of each vehicle and the start and end times of the real-time load restricted section is used as the verification basis. Combined with the parking order relationship, the actual access sequence of the vehicles is checked. The charging behavior that avoids the real-time load restricted section and the access sequence conforms to the bus charging time sorting results is recorded. The charging behavior that falls into the real-time load restricted section and the charging behavior with inconsistent access sequence are marked and summarized to obtain the power grid demand response execution record. The power grid demand response execution record includes actual access records, access that avoids restricted sections, access that falls into restricted sections, and access sequence abnormality records.

[0026] Please see Figure 6 Specifically, in S5, this includes: S51: Based on the vehicle charging sequence in the bus charging time sorting results, read the real-time load change sequence generated by the determined power distribution node during the charging execution phase, organize the real-time load change sequence according to time sequence, call the limit threshold in the demand response command, determine whether the real-time load value has reached the limit threshold, mark the corresponding time period, and obtain the real-time load-limited section; as detailed below. Based on the vehicle charging order in the bus charging time sorting results, i.e., the clearly identified first-priority vehicle number. Second in line and third place The station control data bus retrieves the real-time active power change sequence collected by the transformer-side meters during the corresponding charging execution phase of the distribution node "DN-Bus-Charging-01". The hardware high-frequency sampling period of the retrieval system is set to For each continuously acquired power sample value, digital filtering and numerical analysis are performed, and the previously determined load limit threshold is retrieved from the preset protection parameter register. The power value will be monitored in real time. and Perform a subtraction comparison operation and define a real-time difference variable. If the discriminator determines This indicates that the actual load of the distribution network at that moment has exceeded the safe carrying capacity and is in the over-limit range, for example, during the execution phase. At that moment, the real-time power transmitted back by the meter was Calculate the difference as If the point is determined to be a restricted time, then... If marked as normal, all data will be recorded via a high-precision clock synchronization module. The timestamps of the occurrences are collected and set together. For consecutive restricted time points, the first alarm time of the consecutive segment is extracted as the starting boundary, and the last time normal operation is restored is extracted as the ending boundary. For example, the time from the occurrence of the alarm to the end of the consecutive segment is recorded. to The power values ​​during the period were all within to The price fluctuated wildly between the two, all above the average. If the limiting benchmark is met, then that period is marked as the real-time load high-voltage fluctuation range. The status flag of the demand command issued by the grid side is read synchronously. If the status flag indicates "effective," then the entire time axis coordinates of that segment are mapped to the constrained feature vector space, and the time window is set to [value missing]. The moving average filtering algorithm for seconds performs a logical merge operation, removing items from the set whose durations were less than a certain value due to contactor action. The instantaneous spike pulse exceeding the limit is smoothed out and judged, and only the effective load-limited process with continuous thermodynamic effect is retained to obtain the real-time load-limited section.

[0027] S52: Retrieve the real-time load-limited section, vehicle charging access time, and bus charging time sorting results; read the start and end times of the real-time load-limited section; determine if the vehicle charging access time is outside the section boundary; verify the actual vehicle access order based on the parking sequence; and generate a charging access verification list. Details are as follows: Call the real-time load-constrained section and extract the time boundary. The actual charging access time of each vehicle within the station and the planned access sequence from the bus charging time ranking results are used to extract vehicle information from the charging gun's radio frequency sensing sensor and billing system. The actual physical access time of a successful handshake Retrieve vehicle Actual access time Retrieve vehicle Actual access time Boundary logic is performed one by one according to the vehicle dimension. retrieve its access time Respectively with the starting point of the restricted section and the end point Perform a size comparison and determine if it satisfies the condition of being less than. and less than The condition is that the physical action is completely outside the safe zone bounded by the restricted section boundary. retrieve its access time Determine that it satisfies the condition greater than And greater than The conditions are also outside the restricted section boundary. For each vehicle that has been checked, retrieve the index value of its planned stop order relationship in the sorting schedule. Establish a one-dimensional actual access time sequence array Synchronously retrieve the plan array distributed from the sorting results Perform a logical consistency check on the corresponding positions of array elements, and call the comparison operator to check the index. On Whether or not The string content must be absolutely equal, confirmed by looping through the checks. The vehicle numbers at that time all matched perfectly. If a vehicle was found to be... The actual access time changed due to obstacle avoidance midway through the event. If a time drift occurs but the array order remains second, then the time drift is ignored and the order is considered consistent and valid. However, if the actual access time is incorrect and falls into the wrong range... Within a certain range, if a vehicle violates the instructions... If a forced plug-in is attempted, an error code "Error-01" will be written to the corresponding verification result. If the array order is reversed due to interruption, an error code "Error-02" will be written. A charging access verification list is generated by performing a multi-dimensional and rigorous comparison between the physical access points of all participating vehicles and the sequence array.

[0028] S53: Based on the charging access verification list and real-time load-restricted sections, read the vehicle charging access time, actual access sequence, and bus charging time sorting results; record charging behaviors that avoid real-time load-restricted sections and have the same access sequence; separately mark charging behaviors that fall into real-time load-restricted sections and charging behaviors with inconsistent access sequences; and obtain the grid demand response execution record; as follows: Based on the status flags and specific time ranges of the real-time load-limited sections listed in the charging access verification checklist, the system reads the actual charging access time, actual access location number, and corresponding reference parameters from the bus charging time sorting results for each autonomous driving bus participating in demand response from the server hard drive. For items in the verification checklist with a safety flag of "normal," such as those found... exist It was accessed and is located in the first position of the array; logical verification confirmed that it successfully avoided [the issue]. If the initial power distribution load is constrained during a high-voltage period, and the actual queuing execution order is highly closed with the cloud scheduling plan, then the complete original data packet for this charging behavior is extracted, including the power uploaded from the BMS. The changes, the actual charging power curve uploaded by the charging pile, and the load deviation rate of the master node are stored in the "Compliance Execution" log category table in the database. For items with abnormal codes in the verification list, a violation impact assessment is performed. If a vehicle's actual connection time unfortunately falls into the real-time load-limited section... If "Error-01" is triggered, the pointer is invoked to extract the peak load exceeding the limit on the transformer side at that moment, and the excess load calculation formula is executed. The system accurately records the additional active power impact on the distribution network nodes caused by the illegal charging behavior, and highlights the vehicle's VIN number and the violation timestamp to compile a penalty record. For items reporting "Error-02" (i.e., inconsistent access sequence), the original planned sequence number is retrieved. Compared with the actual serial number monitored ; Calculate displacement deviation by performing absolute subtraction. For example, if it was originally ranked second but actually crossed over to the third position for access, execution... The deviation is calculated as follows: Such disruptive behaviors are separately categorized into the "dispatch deviation" identifier for traceability. By performing structured data cleaning and SQL integration on all compliant, non-compliant, and deviation behaviors throughout the day, a comprehensive evaluation report is generated in the cloud control platform database. This report includes vehicle identification codes, time period response success rates, peak reduction amount contributed by distribution side loads, and quantified statistics of time deviation within the station, thus obtaining the power grid demand response execution record.

[0029] A power grid demand response system based on dynamic optimization of autonomous driving bus driving strategies, comprising: The trajectory filtering module collects the vehicle speed sequence, real-time location, predicted arrival time and battery status of autonomous buses, reads the planned arrival time, identifies stable driving sections and trajectory sections, marks the time offset status, determines the arrival time adjustment interval, verifies the battery availability status, and filters out a set of trajectories that can participate in demand response. The load identification module determines the distribution node corresponding to the grid connection location of the target charging station, obtains the load change sequence and sampling time, reads the demand response command, identifies continuous load-limited sections according to the sampling time, compares them with the overlapping load-limited time periods, merges the continuous overlapping time periods, and obtains the grid load-limited time period. The access classification module extracts the predicted arrival time and arrival time adjustment interval from the set of available demand response trajectories, reads the start and end times of the grid load restriction time segment, marks the three access types as early, late, and sequential according to the time position relationship, and sorts them to obtain the charging access timing adjustment set. The rhythm control module determines the vehicle driving and stopping adjustment mode based on the access type mark of the charging access timing adjustment set. Through vehicle control commands, it controls the vehicle to move forward, delay, or maintain the original rhythm. Based on the adjusted arrival time, access time, and stopping order, it obtains the bus charging time sorting result. The execution verification module reads the real-time load change sequence of the determined power distribution nodes based on the bus charging time sorting results, marks the real-time load-restricted sections, verifies the vehicle charging access time and actual access sequence, records compliant charging behavior, summarizes abnormal charging behavior, and obtains the power grid demand response execution record.

[0030] It is understandable that the above system and method have the same execution process and the same effect, so they will not be described again here.

[0031] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy, characterized in that, Includes the following steps: S1: Collect the vehicle speed sequence, real-time location, predicted arrival time and battery status of autonomous buses, read the planned arrival time, identify stable driving sections and trajectory sections, mark the time offset status, determine the arrival time adjustment interval, verify the battery availability status, and filter to obtain the set of trajectories that can participate in demand response. S2: Determine the distribution node corresponding to the grid access location of the target charging station, obtain the load change sequence and sampling time, read the demand response command, identify the continuous load-limited section according to the sampling time, compare it with the load-limited time period, merge the continuous overlapping time periods, and obtain the grid load-limited time period. S3: Extract the predicted arrival time and arrival time adjustment interval from the set of available demand response trajectories, read the start and end times of the power grid load restriction time segment, mark the three access types of early, late and sequential according to the time position relationship, and sort them to obtain the charging access timing adjustment set. S4: Based on the access type marker of the charging access timing adjustment set, determine the vehicle driving and stopping adjustment mode, and control the vehicle driving forward, backward or maintain the original rhythm through vehicle control commands. According to the adjusted arrival time, access time and stopping order, obtain the bus charging time sorting result.

2. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 1, characterized in that: The set of trajectories eligible for demand response includes trajectory segments, adjustable arrival time ranges, available power status, and time offset status. The grid load restriction time segments include continuous load-limited segments, overlapping load restriction segments, and merged continuous segments. The set of charging access timing adjustment includes early access trajectories, delayed access trajectories, sequential access trajectories, and trajectories arranged according to predicted arrival order. The bus charging time sorting results include adjusted predicted arrival time, charging access time, vehicle charging sequence, and stop order.

3. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 1, characterized in that: The steps for obtaining S1 are as follows: S11: Collect the speed change sequence of autonomous buses, real-time vehicle location, predicted arrival time and vehicle battery status, read the planned arrival time in the bus operation plan, organize the speed change sequence according to the operation sequence, check whether the speed fluctuation is within the preset range, delineate the driving trajectory segment by associating the real-time vehicle location, and obtain the stable trajectory segment. S12: Call the stable trajectory segment, predicted arrival time and planned arrival time, mark the early, on-time and late status according to the arrival order, and determine the allowable forward and backward movement boundaries of the predicted arrival time in combination with the continuous status of the stable driving segment, and generate the arrival time adjustment interval; S13: Based on the arrival time adjustment interval and vehicle battery status, read the minimum power requirement of the target charging station, verify the charging station access requirements corresponding to the remaining power, mark the power availability status, filter the trajectory segments with arrival time adjustment interval and power availability status that meet the requirements, and obtain the set of trajectories that can participate in demand response.

4. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 1, characterized in that: The steps for obtaining S2 are as follows: S21: Determine the corresponding distribution node based on the grid access location of the target charging station, obtain the load change sequence and corresponding sampling time output by the load monitoring equipment of the distribution node according to the preset sampling period, read the load restriction period and load restriction threshold in the demand response command issued by the grid side, organize the load value according to the sampling time, compare the sampling time boundary with the load restriction period boundary, and obtain the node load reference time sequence; S22: Call the load change sequence, corresponding sampling time and load limit threshold in the node load reference time series, read the paired load values ​​according to adjacent sampling time, compare the paired load values ​​with the load limit threshold respectively, select the time period when both load values ​​reach the load limit threshold, compare the boundary overlap with the load limit time period, and obtain the load limit overlap section; S23: Based on the load limiting overlapping section, the preset sampling period and the corresponding sampling time, arrange the overlapping time periods in the order of the start time, compare the interval between the end time of adjacent overlapping time periods and the next start time, merge the overlapping time periods that are continuous in time and whose interval does not exceed the preset sampling period, and record the start time and end time after merging to obtain the power grid load limiting time section.

5. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 1, characterized in that: The steps for obtaining S3 are as follows: S31: Extract the predicted arrival time and arrival time adjustment interval corresponding to each trajectory segment in the set of trajectories that can participate in demand response trajectories, read the start time and end time of each power grid load restriction time segment, associate the predicted arrival time, adjustment interval boundary and load restriction boundary according to the trajectory segment number, compare the order relationship between the predicted arrival time and the load restriction boundary, and obtain the access location relationship value. S32: Call the access location relationship value, predicted arrival time, arrival time adjustment interval end time, load limit start time and load limit end time, determine the three types of relationships: the predicted arrival time is within the load limit interval, earlier than the load limit start time, and later than the load limit end time, and mark the delayed access type, early access type and sequential access type respectively, and generate access type tags; S33: Based on the access type marker and the access location relationship value, read the corresponding trajectory segments of delayed access type, early access type and sequential access type, organize the trajectory segments of each type in the order of predicted arrival time, check the correspondence between trajectory segment number, arrival time adjustment interval and power grid load restriction time segment, and obtain the charging access timing adjustment set.

6. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 5, characterized in that: The specific steps for organizing various types of trajectory segments according to the predicted arrival times are as follows: According to the access type marker, the trajectory segments corresponding to the delayed access type, the early access type, and the sequential access type are read respectively, and the predicted arrival time, the arrival time adjustment interval, and the power grid load restriction time interval corresponding to the trajectory segment are kept in association. Based on the order of the predicted arrival time, an intra-type arrangement record is formed respectively. The specific correspondence between the trajectory segment number, arrival time adjustment interval, and power grid load restriction time segment is as follows: The trajectory segment number is matched with the arrival time adjustment interval, the power grid load restriction time interval, and the access location relationship value one by one. The trajectory segments whose access type marker matches the access location relationship value are retained, and the retained trajectory segments are written into the charging access timing adjustment set.

7. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 1, characterized in that: The steps for obtaining S4 are as follows: S41: Based on the access type marker of each trajectory segment in the charging access timing adjustment set, read the trajectory segment number, predicted arrival time, arrival time adjustment interval and grid load limitation time segment, determine the state of driving rhythm forward, stopping rhythm delayed or original rhythm maintained according to the access type marker, call the vehicle driving controller to write the corresponding control content, and generate vehicle control instructions. S42: Call the vehicle control command, arrival time adjustment interval, start time and end time of the power grid load restriction time segment, adjust the predicted arrival time to be within the arrival time adjustment interval and not later than the start time of the corresponding power grid load restriction time segment for the early access type, adjust the charging access time to be not earlier than the end time of the corresponding power grid load restriction time segment for the delayed access type, and maintain the original driving and stopping rhythm for the sequential access type to obtain the access time boundary interval; S43: Based on the access time boundary interval, vehicle control command and sequential access type trajectory segment, read the adjusted predicted arrival time, charging access time and stopping order relationship, arrange the charging order of vehicles with early access type, delayed access type and sequential access type according to the time sequence, and obtain the bus charging time sorting result.

8. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 1, characterized in that: The method further includes: S5: Based on the bus charging time sorting results, read the real-time load change sequence of the determined power distribution nodes, mark the real-time load-limited sections, verify the vehicle charging access time and actual access sequence, record compliant charging behavior, summarize abnormal charging behavior, and obtain the power grid demand response execution record. The power grid demand response execution record includes actual access records, access that avoids restricted sections, access that falls into restricted sections, and access sequence abnormality records.

9. The power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy according to claim 8, characterized in that: The steps for obtaining S5 are as follows: S51: Based on the vehicle charging sequence in the bus charging time sorting result, read the real-time load change sequence generated by the determined power distribution node during the charging execution phase, organize the real-time load change sequence according to the time sequence, call the limit threshold in the demand response instruction, determine whether the real-time load value has reached the limit threshold, mark the corresponding time period, and obtain the real-time load-limited section. S52: Call the sorting results of the real-time load-limited section, the charging access time of each vehicle and the bus charging time, read the start time and end time of the real-time load-limited section, determine whether the vehicle charging access time is outside the section boundary, and verify the actual access order of the vehicles in combination with the parking order relationship to generate a charging access verification list. S53: Based on the charging access verification list and the real-time load-limited section, read the vehicle charging access time, actual access order and bus charging time sorting results, record the charging behavior that avoids the real-time load-limited section and has the same access order, and mark the charging behavior that falls into the real-time load-limited section and has an inconsistent access order, to obtain the grid demand response execution record.

10. A power grid demand response system based on dynamic optimization of autonomous driving bus driving strategies, characterized in that, The system is used in the power grid demand response method based on dynamic optimization of autonomous driving bus driving strategy as described in any one of claims 1-9, and the system comprises: The trajectory filtering module collects the vehicle speed sequence, real-time location, predicted arrival time and battery status of autonomous buses, reads the planned arrival time, identifies stable driving sections and trajectory sections, marks the time offset status, determines the arrival time adjustment interval, verifies the battery availability status, and filters out a set of trajectories that can participate in demand response. The load identification module determines the distribution node corresponding to the grid connection location of the target charging station, obtains the load change sequence and sampling time, reads the demand response command, identifies continuous load-limited sections according to the sampling time, compares them with the overlapping load-limited time periods, merges the continuous overlapping time periods, and obtains the grid load-limited time period. The access classification module extracts the predicted arrival time and arrival time adjustment interval from the set of available demand response trajectories, reads the start and end times of the power grid load restriction time segment, marks the three access types as early, late, and sequential according to the time position relationship, and sorts them to obtain the charging access timing adjustment set. The rhythm control module determines the vehicle driving and stopping adjustment mode based on the access type mark of the charging access timing adjustment set. Through vehicle control commands, it controls the vehicle to move forward, delay, or maintain the original rhythm. Based on the adjusted arrival time, access time, and stopping order, it obtains the bus charging time sorting result. The execution verification module reads the real-time load change sequence of the determined power distribution nodes based on the bus charging time sorting results, marks the real-time load-limited sections, verifies the vehicle charging access time and actual access sequence, records compliant charging behavior, summarizes abnormal charging behavior, and obtains the power grid demand response execution record.

Images