Bus shift extraction optimization method based on multi-dimensional score driving

Abstract

The invention discloses a bus shift extraction optimization method based on multi-dimensional score driving, and relates to the field of urban public traffic scheduling management, and the method comprises the following steps: initializing shift extraction parameters, determining a shift extraction target, defining time period features, and setting a score weight and a penalty coefficient; constructing a shift extraction optimization mathematical model, including defining decision variables and establishing an objective function and constraint conditions thereof; respectively constructing a multi-dimensional scoring system and a constraint verification strategy by utilizing the objective function and the constraint conditions thereof; generating candidate shift drawing schemes, calculating a comprehensive score and selecting an optimal shift drawing scheme; performing iteration processing on the optimal shift extraction scheme in combination with loop control and a scheme rollback strategy to obtain a final shift extraction scheme; and verifying and counting the final shift drawing scheme, and generating a corresponding shift drawing report. On the premise that operation balance and service integrity are ensured, global optimal selection of the shift drawing scheme is achieved, and the intelligent level and operation efficiency of bus dispatching are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of urban public transportation dispatch management, and more specifically to a bus shift optimization method based on multi-dimensional scoring. Background Technology

[0002] In urban public transportation dispatch and management, optimizing bus schedules is a key aspect of improving operational efficiency and service quality. One common practice is to reduce double-shift buses to single-shift buses (shift cancellation), which requires scientifically and rationally reducing the number of buses while ensuring operational continuity and directional balance, in order to adapt to changes in passenger flow and resource allocation needs.

[0003] Existing shift scheduling methods often lack a systematic multi-objective collaborative optimization mechanism. Traditional methods rely heavily on single factors for decision-making, failing to comprehensively consider multiple dimensions such as peak-hour impact, passenger flow matching, and operational stability. As a result, local optimization leads to global imbalance, making it difficult to generate shift scheduling plans that perform well across multiple operational indicators, resulting in insufficient scientific rigor and adaptability of the overall scheduling strategy.

[0004] A further drawback is that existing technologies are weak in handling complex business constraints. The bus dispatching process is subject to strict constraints such as continuity, directionality, quantity, and completeness. When faced with conflicting constraints, existing methods lack effective verification and reconciliation mechanisms, which can easily lead to invalid solutions that violate operational rules or severely sacrifice other operational requirements to meet one constraint, affecting the feasibility of the solution in a real system.

[0005] Furthermore, existing methods have significant shortcomings in terms of the flexibility and control precision of shift scheduling strategies. On the one hand, they fail to intelligently identify and utilize the characteristic value of different patterns, such as splitting shift scheduling; on the other hand, they cannot independently and precisely control the number of shifts in the up and down directions. This leads to a loss of directional balance in route operations, resulting in problems such as uneven distribution of vehicle resources and a mismatch between service capacity and passenger demand during certain periods.

[0006] Therefore, how to design a bus shift optimization method based on multi-dimensional scoring to support the bus dispatching system in achieving efficient, balanced and reliable shift adjustments is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] In view of this, the present invention provides a bus shift selection optimization method based on multi-dimensional scoring, which aims to solve the problems of existing bus shift selection methods such as lack of systematic evaluation, imperfect scoring mechanism, improper handling of constraint conflicts, and insufficient directional control, so as to realize the automation and intelligence of efficient and reasonable conversion from double shift buses to single shift buses, and ultimately ensure the stability, balance and efficiency of bus operation.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A bus shift optimization method based on multi-dimensional scoring includes the following steps:

[0010] S1. Initialize the sampling parameters, determine the sampling target, define the time period characteristics, and set the scoring weights and penalty coefficients;

[0011] S2. Based on the sampling parameters, construct a sampling optimization mathematical model, including defining decision variables, establishing the objective function and its constraints;

[0012] S3. Using the objective function and its constraints, construct a multi-dimensional scoring system and a constraint verification strategy respectively;

[0013] S4. Based on the multi-dimensional scoring system and constraint verification strategy, generate candidate class selection schemes, calculate the comprehensive score, and select the optimal class selection scheme.

[0014] S5. Combining loop control and scheme rollback strategy, the optimal shift removal scheme is iteratively processed to obtain the final shift removal scheme.

[0015] S6. Verify and statistically analyze the final shift selection plan, and generate the corresponding shift selection report.

[0016] Preferably, in S1, the shift allocation target includes the up-line single-shift vehicle allocation limit obtained from the main station and the down-line single-shift vehicle allocation limit obtained from the secondary station;

[0017] The time period characteristics include the defined morning rush hour and evening rush hour;

[0018] The scoring weights and penalty coefficients include: peak non-overlap weight α1, departure interval weight α2, mode feature weight β, vehicle number weight γ, passenger flow matching weight δ, operational stability weight ε, and first and last bus penalty coefficients; the mode feature weight β includes splitting and randomizing the shift weight mode β1 and last bus mode β2; the passenger flow matching weight includes density matching weight δ1, load factor weight δ2, and capacity balance weight δ3.

[0019] Preferably, in step S2, the decision variables are defined as follows:

[0020] Define a vehicle set V = {v1, v2, ..., v...} n}, where n is the total number of vehicles;

[0021] For vehicle v i Define its class set m i For vehicle v i The total number of flights;

[0022] Define the decision variable x for random shifts ij x ∈{0,1} ij =1 indicates that the first [value] is retained. Each class, x ij =0 indicates that the shift has been deleted;

[0023] Define direction indicator variable d i ∈{0,1},d i =0 indicates the upward direction, d i =1 indicates the downward direction.

[0024] Preferably, in step S2, establishing the objective function maxF(x) includes:

[0025]

[0026] Where, f1(x) ij f2(x) ij f3(x) ij f4(x) ij f5(x) ij f6(x) ij ) represent the scoring functions for peak hours, departure intervals, mode characteristics, vehicle number, passenger flow matching, and operational stability, respectively, p(x ij ) represents the penalty function, and w1, w2, w3, w4, w5, and w6 represent the weight coefficients for each dimension.

[0027] Preferably, in S2, the constraints include:

[0028] The continuity constraint is used to limit the continuity of the remaining shifts after shift selection. It requires that the number of consecutive segments does not exceed 2 and that the parity rule is satisfied.

[0029] Directional constraints require that the direction of the earliest departure event of a vehicle after the shift is drawn must be consistent with the initial direction of the vehicle.

[0030] Quantity constraints limit the total number of shifts drawn in both the upward and downward directions to a preset limit.

[0031] The integrity constraint requires that the number of shifts retained for each vehicle after shift selection is no less than the minimum required number of shifts.

[0032] Preferably, in S3, the multi-dimensional scoring system includes:

[0033] Peak hour scoring is used to assess the impact of shift scheduling on peak hour operations;

[0034] Departure interval score is used to evaluate the uniformity and rationality of departure intervals after random shifts are taken.

[0035] Pattern feature scoring is used to identify and differentiate patterns such as split shifts, consecutive shifts, and those involving the last bus.

[0036] Vehicle number scoring is used to assign specific bonus points based on the vehicle number to aid in decision-making.

[0037] Passenger flow matching score is used to evaluate the degree of matching between the shift scheduling plan and passenger flow demand, including passenger flow density, load factor and capacity balance;

[0038] Operational stability score is used to assess the impact of shift changes on vehicle utilization and operational stability;

[0039] The first and last bus penalty score is used to impose severe penalties on schemes that remove the first or last bus.

[0040] Preferably, in S3, the constraint verification strategy includes:

[0041] Continuity check: Verify whether the remaining shift index forms a valid continuous segment and meets the parity requirement;

[0042] Directional verification: Verify whether the direction of the earliest departure event after the shift selection is consistent with the initial direction of the vehicle.

[0043] Completeness check to verify whether the plan after shift sampling meets the minimum number of shifts requirement and the rationality of the operation trajectory.

[0044] Preferably, S4 includes:

[0045] S41. Traverse the vehicle set, filter the vehicles to be evaluated based on directional constraints, calculate the target number of retained shifts based on the original number of shifts of the vehicles, generate candidate shift deletion combinations using a combination algorithm, and apply constraint verification strategy to filter out feasible candidate solutions.

[0046] S42. For each feasible candidate solution, call the various scoring functions in the multi-dimensional scoring system to calculate its score, and combine them with preset weights to perform weighted summation to obtain the comprehensive score of the solution.

[0047] S43. Sort all feasible candidate solutions in descending order of comprehensive score, and select the solution ranked first as the current optimal solution for class selection.

[0048] S44. Based on the selected optimal solution, delete the specified shift from the corresponding vehicle shift table, update the vehicle status to single shift, and update the single shift vehicle allocation counter for the corresponding direction.

[0049] Preferably, S5 includes:

[0050] S51. At the beginning of each iteration, determine whether the current total number of shifts drawn has reached the preset shift drawing target, and determine whether the number of shifts drawn in the upward and downward directions has reached their preset limits respectively; if any condition is met, terminate the iteration process.

[0051] S52. When the current round of shift selection fails because a feasible solution cannot be generated or an invalid solution is generated, the vehicle shift schedule and related counters will be rolled back to the state before the current iteration.

[0052] S53. After generating the candidate optimal shift selection scheme in each iteration, verify whether the scheme satisfies all mathematical constraints and business rules, and confirm the correctness of the vehicle status update.

[0053] S54. For the verified scheme, update the system status, including the total number of shifts, the shift counters for each direction, and the vehicle status.

[0054] Preferably, in step S6, the shift report includes:

[0055] A summary of the shift change plan, listing all vehicles whose shifts were changed and the specific shifts that were deleted;

[0056] The scoring analysis report shows the final solution's scoring details across various dimensions;

[0057] A constraint satisfaction report, describing the satisfaction status of all mathematical and business constraints;

[0058] Optimize the effect evaluation and provide comparative analysis of key operational indicators before and after shift changes.

[0059] As can be seen from the above technical solution, compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0060] 1. This method transforms the original human experience-dependent shift selection decision into a quantifiable comprehensive evaluation and optimization problem by constructing a multi-dimensional scoring system and mathematical optimization model. This significantly improves the scientific nature and global optimality of the shift selection scheme and overcomes the decision-making limitations caused by the one-sided reliance on a single factor in traditional methods.

[0061] 2. By introducing a multi-constraint verification mechanism including continuity, directionality, quantity, and completeness, operational conflicts can be systematically identified and avoided during the shift selection process. This ensures that the generated shift selection plan conforms to both mathematical constraints and actual business rules, thereby improving the feasibility and reliability of the plan in actual operation.

[0062] 3. By setting limits on the number of buses to be drawn in both directions and combining this with verification of the initial direction consistency of vehicles, precise control over the operating direction is achieved, avoiding route operation imbalances caused by uneven bus draws and improving the overall coordination of public transport services and passenger satisfaction. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0064] Figure 1 A flowchart of a bus shift optimization method based on multi-dimensional scoring is provided for an embodiment of the present invention.

[0065] Figure 2 This is a schematic diagram illustrating the process of selecting the optimal shift selection scheme provided in an embodiment of the present invention;

[0066] Figure 3 This is a schematic diagram illustrating the process of obtaining the final shift selection scheme provided in an embodiment of the present invention. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] like Figure 1 As shown, this embodiment provides a bus shift optimization method based on multi-dimensional scoring, including the following steps:

[0069] S1. Initialize the sampling parameters, determine the sampling target, define the time period characteristics, and set the scoring weights and penalty coefficients;

[0070] S2. Based on the sampling parameters, construct a sampling optimization mathematical model, including defining decision variables, establishing the objective function and its constraints;

[0071] S3. Using the objective function and its constraints, construct a multi-dimensional scoring system and a constraint verification strategy respectively;

[0072] S4. Based on the multi-dimensional scoring system and constraint verification strategy, generate candidate class selection schemes, calculate the comprehensive score, and select the optimal class selection scheme.

[0073] S5. Combining loop control and scheme rollback strategy, the optimal shift removal scheme is iteratively processed to obtain the final shift removal scheme.

[0074] S6. Verify and statistically analyze the final shift selection plan, and generate the corresponding shift selection report.

[0075] By establishing an optimization method that combines a multi-dimensional scoring system with intelligent constraint verification, the bus dispatching process is transformed from relying on human experience to a systematic and quantitative scientific decision-making process. This effectively solves scheduling conflicts under multiple constraints and achieves the global optimal selection of dispatching schemes while ensuring operational balance and service integrity. This significantly improves the intelligence level and operational efficiency of bus dispatching.

[0076] The following provides a further detailed explanation of each step in the above method;

[0077] In this embodiment, S1 is to initialize the sampling parameters, determine the sampling target, define the time period characteristics, and set the scoring weight and penalty coefficient.

[0078] The train allocation targets include the number of trains allocated per shift for the uphill route obtained from the main station and the number of trains allocated per shift for the downhill route obtained from the secondary station.

[0079] The time period characteristics include the defined morning peak period of 07:00-09:00 and the evening peak period of 17:00-19:00;

[0080] The scoring weights and penalty coefficients include: peak non-overlap weight α1=40, departure interval weight α2=100, mode feature weight β, vehicle number weight γ=5, passenger flow matching weight δ, operational stability weight ε, and a penalty coefficient of -10000 for the first and last buses; the mode feature weight β includes splitting and randomizing the shift weight mode β1=40 and the last bus mode β2=100; the passenger flow matching weight includes density matching weight δ1=80, load factor weight δ2=75, and capacity balance weight δ3=55;

[0081] This step establishes a quantitative benchmark that balances operational needs and optimization objectives by precisely setting limits on the number of vehicles assigned per shift for both upbound and downbound routes, clarifying peak hours, and configuring a complete parameter system that includes weights for peak non-overlap, departure intervals, mode characteristics, vehicle numbers, passenger flow matching, operational stability, and first and last bus penalties. This provides a comprehensive and adjustable parameter foundation for subsequent mathematical modeling and intelligent decision-making.

[0082] In this embodiment, S2, based on the sampling parameters, constructs a sampling optimization mathematical model, including defining decision variables, establishing an objective function and its constraints;

[0083] The decision variables are defined as follows:

[0084] Define a vehicle set V = {v1, v2, ..., v...} n}, where n is the total number of vehicles;

[0085] For vehicle v iDefine its class set m i For vehicle v i The total number of flights;

[0086] Define the decision variable x for random shifts ij x ∈{0,1} ij =1 indicates that the first [value] is retained. Each class, x ij =0 indicates that the shift has been deleted;

[0087] Define direction indicator variable d i ∈{0,1},d i =0 indicates the upward direction, d i =1 indicates the downward direction.

[0088] Furthermore, establishing the objective function maxF(x) includes:

[0089]

[0090] Where, f1(x) ij f2(x) ij f3(x) ij f4(x) ij f5(x) ij f6(x) ij ) represent the scoring functions for peak hours, departure intervals, mode characteristics, vehicle number, passenger flow matching, and operational stability, respectively, p(x ij ) represents the penalty function, and w1, w2, w3, w4, w5, and w6 represent the weight coefficients for each dimension;

[0091] Specifically, f1(x) ij ) is the scoring function for peak hours, f1(x) ij )=(N total -N peak )×α1,N total N represents the total number of classes sampled. peak f2(x) represents the number of shifts sampled during peak hours. ij f2(x) is the departure interval scoring function. ij f3(x) = min(α2 / avg_interval, α2 / min_interval), where avg_interval is the average occurrence interval and min_interval is the minimum occurrence interval; ij f3(x) is the pattern feature scoring function. ij )=β1·I split +β2·I last I split To split the shift indicator variable, I lastf4(x) is the indicator variable for the last train. ij f4(x) is the vehicle number scoring function. ij )=γ·vehicle_id;f5(x ij f5(x) is the passenger flow matching scoring function. ij )=δ1·M density +δ2·M load +δ3·M balance M density M load M balance These are passenger flow density matching index, passenger flow load matching index, and passenger flow balance matching index, respectively; f6(x ij f6(x) is the operational stability scoring function. ij )=ɛ·U utilization ;p(x ij Let p(x) be the penalty function. ij )=-10000·(I first +I last ), I first The first train is indicated by this variable.

[0092] Furthermore, the constraints include:

[0093] Continuity constraints are used to limit the continuity of remaining shifts after shift cancellation, requiring that the number of consecutive segments does not exceed 2 and that the parity rule is satisfied; for any vehicle v i The remaining train services must meet the continuity requirement:

[0094] , For the target class number; , indicating that the number of consecutive segments does not exceed 2; , ; , then (first_segment_end ten second_segment_start)mod 2 = 1;

[0095] Directional constraints require that the earliest departure event direction of a vehicle after the shift is drawn must be consistent with the initial direction of the vehicle; direction(first_remaining_trip(v i )) = d i , where first_remaining_trip(v i ) for vehicle v i The first shift after the shift assignments were completed;

[0096] Quantity constraints limit the total number of shifts drawn in both the upward and downward directions to a preset limit. There is a limit to the number of shifts that can be drawn in the upward direction; The number of shifts that can be withdrawn in the downward direction is limited;

[0097] The integrity constraint requires that the number of shifts retained for each vehicle after shift removal is not less than the minimum required number of shifts;

[0098] ,in The minimum number of required shifts.

[0099] This section constructs a complete mathematical representation system by defining vehicle sets, shift sets, shift selection decision variables, and direction indicator variables. Furthermore, the established multi-dimensional objective function integrates six scoring dimensions and a penalty mechanism, along with four types of strict mathematical constraints: continuity constraints, directional constraints, quantity constraints, and integrity constraints. This transforms complex business problems into solvable 0-1 integer programming models, and transforms complex shift selection problems into optimization problems with clear decision variables, multi-dimensional objective functions, and strict constraints, providing a reliable mathematical framework and solution foundation for subsequent intelligent decision-making.

[0100] In this embodiment, S3, a multi-dimensional scoring system and a constraint verification strategy are constructed using the objective function and its constraints, respectively.

[0101] The multi-dimensional scoring system includes:

[0102] Peak hour scoring is used to assess the impact of shift change plans on peak hour operations; this includes calculating the impact of shift change plans on peak hours.

[0103] 1) Define the set of peak time periods:

[0104] 2) Calculate the number of shifts to be drawn during peak hours:

[0105] 3) Non-overlapping peak scores:

[0106] ,in ;

[0107] The departure interval score is used to evaluate the uniformity and rationality of the departure interval after shift changes; it includes evaluating the impact of shift changes on the departure interval.

[0108] 1) Calculate the average departure interval after shift changes: ;

[0109] 2) Calculate the minimum departure interval after shift changes: 3) Departure interval score: ;

[0110] Pattern feature scoring is used to identify and differentiate patterns of split-shift, consecutive-shift, and shift-related patterns involving the last bus; this includes identifying and evaluating shift patterns:

[0111] 1) Continuous shift selection mode: Otherwise, it is 0; 2) Split the shift selection mode: If the remaining buses after the shift change can be divided into exactly two consecutive segments, otherwise the value is 0; 3) Last bus mode: 4) Pattern feature score: (If it's the last train, then it's 0;) ;

[0112] Vehicle identification number (VIN) scoring is used to assign specific bonus points based on the vehicle ID to aid decision-making; this includes bonus points based on the vehicle ID:

[0113] 1) Vehicle number score: ,in Number the vehicle.

[0114] Passenger flow matching score is used to evaluate the degree of matching between the shift change plan and passenger flow demand, including passenger flow density, load factor, and capacity balance; it also includes an assessment of the impact of shift changes on passenger flow matching.

[0115] 1) Passenger flow density matching score: ;in,

[0116] For departure density, 1) Passenger flow density; 2) Load factor optimization score: ;in, The target load factor is 70-80%. 3) Score for balanced capacity distribution: (Actual load factor) ;in, For capacity standard deviation, 4) Average capacity; Total passenger flow matching score: ;

[0117] Operational stability score, used to assess the impact of shift changes on vehicle utilization and operational stability; including assessing the impact of shift changes on operational stability:

[0118] 1) Vehicle utilization rate: ,in This refers to the actual operating time.

[0119] Total available time; 2) Utilization improvement score: ;in, The baseline utilization rate is 85-90%.

[0120] The first and last bus penalty score is used to impose severe penalties on schemes that remove the first or last bus; this includes penalties for schemes involving the first and last buses.

[0121] 1) Penalty for the first bus: ,in

[0122] 1) First bus, otherwise 0; 2) Penalty for last bus: 3) Total penalty: ;

[0123] Furthermore, constraint verification strategies include:

[0124] The continuity check verifies whether the remaining shift indexes form valid continuous segments and meet the parity requirements; including: 1) grouping the remaining indexes by continuity to form continuous segments; 2) checking whether the number of continuous segments does not exceed 2; 3) if it is a single continuous segment, the number of continuous segments must be even; 4) if it is 2 continuous segments, the last index of the first continuous segment and the first index of the second continuous segment must be parity.

[0125] Directional verification verifies whether the direction of the earliest departure event after the shift selection is consistent with the initial direction of the vehicle; including 1) obtaining the earliest index in the remaining index; 2) checking whether the direction of the earliest departure event is consistent with the initial direction of the vehicle; 3) if they are inconsistent, marking the combination as invalid.

[0126] Integrity check: Verify whether the plan after shift selection meets the minimum number of shifts requirement and the rationality of the operation trajectory; including 1) checking whether enough shifts have been retained; 2) verifying the rationality of the vehicle operation trajectory; 3) ensuring that the vehicles can complete the operation tasks normally.

[0127] The scoring system built in this step fully covers seven components: peak hour scoring, departure interval scoring, mode characteristic scoring, vehicle number scoring, passenger flow matching scoring, operational stability scoring, and first and last bus penalty scoring. Together with the constraint verification strategy, which includes continuity checks, directional verification, and integrity checks, it forms a dual guarantee mechanism for scheme evaluation and screening, ensuring that each candidate scheme not only performs well in multiple operational indicators, but also meets various hard constraints in actual business.

[0128] In this embodiment, S4 involves generating candidate class selection schemes based on the multi-dimensional scoring system and constraint verification strategy, calculating the comprehensive score, and selecting the optimal class selection scheme; as follows: Figure 2 As shown, it specifically includes:

[0129] S41. Traverse the vehicle set, filter vehicles to be evaluated based on directional constraints, calculate the target number of retained shifts based on the original number of shifts for each vehicle, generate candidate shift deletion combinations using a combination algorithm, and apply constraint verification strategies to select feasible candidate solutions; where the target number of retained shifts... Represented as:

[0130]

[0131] The specific combination algorithm is as follows: using a recursive backtracking method or an enumeration algorithm based on binomial coefficients, all possible combinations of deleting k shifts from the vehicle shift set are traversed, where k is the number of shifts to be deleted; during the traversal process, the constraint verification strategy is applied first for pre-screening, and early pruning is performed on some combinations that violate continuity or direction constraints, so as to reduce the number of candidate solutions that need to be fully evaluated and improve the efficiency of generating feasible candidate solutions;

[0132] S42. For each feasible candidate solution, call the various scoring functions in the multi-dimensional scoring system to calculate its score, and combine them with preset weights to perform weighted summation to obtain the comprehensive score of the solution.

[0133] S43. Sort all feasible candidate solutions in descending order of comprehensive score, and select the solution ranked first as the current optimal solution for class selection.

[0134] S44. Based on the selected optimal solution, delete the specified shift from the corresponding vehicle shift table, update the vehicle status to single shift, and update the single shift vehicle allocation counter for the corresponding direction.

[0135] This step filters vehicles through directional constraints, calculates the target number of shifts, and uses a combination algorithm of recursive backtracking or binomial enumeration to generate candidate deletion combinations and apply constraints for pre-screening, which significantly improves generation efficiency. Then, it realizes multi-dimensional quantitative evaluation through comprehensive score calculation, sorts the scores in descending order and selects the optimal solution, and finally executes shift deletion and updates vehicle status and direction counters, realizing efficient and intelligent selection of the current optimal solution from a massive number of possible solutions.

[0136] In this embodiment, S5, combining loop control and scheme rollback strategy, iteratively processes the optimal shift selection scheme to obtain the final shift selection scheme; as follows: Figure 3 As shown, it specifically includes:

[0137] S51. At the beginning of each iteration, determine whether the current total number of shifts drawn has reached the preset shift drawing target, and determine whether the number of shifts drawn in the upward and downward directions has reached their preset limits respectively; if any condition is met, terminate the iteration process.

[0138] S52. When the current round of shift selection fails because a feasible solution cannot be generated or an invalid solution is generated, the vehicle shift schedule and related counters will be rolled back to the state before the current iteration.

[0139] S53. After generating the candidate optimal shift selection scheme in each iteration, verify whether the scheme satisfies all mathematical constraints and business rules, and confirm the correctness of vehicle status updates; including checking whether the number of shifts meets the requirements: verifying whether all constraints are met; verifying whether the number of shifts in each direction is balanced: checking the satisfaction of directional constraints; confirming that all vehicle statuses are correct: verifying the satisfaction of integrity constraints.

[0140] S54. For the verified scheme, update the system status, including the total number of shifts, the shift counters for each direction, and the vehicle status.

[0141] This iterative optimization mechanism, through loop control, state rollback, and result verification, ensures that the sampling process can stably and controllably approach the global optimal goal, thereby improving the system's robustness and reliability in dealing with complex constraints and uncertainties.

[0142] In this embodiment, S6 involves verifying and statistically analyzing the final shift selection plan, and generating a corresponding shift selection report; specifically, it includes:

[0143] The report includes a summary of shift change plans, listing all vehicles affected by the change and the specific shifts that were removed; a scoring analysis report, showing the final plan's scoring details across various dimensions; a constraint satisfaction report, explaining the satisfaction status of all mathematical and business constraints; and an optimization effect evaluation, providing a comparative analysis of key operational indicators before and after the shift change.

[0144] This step generates a summary of shift selection plans, a scoring analysis report, a constraint satisfaction report, and an optimization effect evaluation. It not only provides an executable shift selection plan, but also builds a comprehensive output that covers plan interpretation, effect verification, and decision support, thus achieving transparency, auditability, and continuous optimization capabilities in the optimization process.

[0145] This embodiment presents a bus shift scheduling optimization method based on multi-dimensional scoring. Using a precise parameter system, it transforms the shift scheduling problem into a quantifiable 0-1 integer programming model. The method evaluates solutions based on a multi-dimensional scoring system and ensures business compliance through strict constraint verification strategies. It enhances process robustness through iterative optimization and rollback mechanisms, ultimately outputting a feasible and optimal shift scheduling solution and a complete report. This method effectively overcomes the limitations of traditional reliance on manual experience, significantly improving aspects such as balancing up and down-line operations, ensuring peak capacity, matching passenger flow demand, and stabilizing operational efficiency. It provides a reliable practical solution for the transformation of bus scheduling from experience-based to intelligent methods.

[0146] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0147] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A bus shift selection optimization method based on multi-dimensional scoring, characterized in that, Includes the following steps: S1. Initialize the sampling parameters, determine the sampling target, define the time period characteristics, and set the scoring weights and penalty coefficients; S2. Based on the sampling parameters, construct a sampling optimization mathematical model, including defining decision variables, establishing the objective function and its constraints; S3. Using the objective function and its constraints, construct a multi-dimensional scoring system and a constraint verification strategy respectively; S4. Based on the multi-dimensional scoring system and constraint verification strategy, generate candidate class selection schemes, calculate the comprehensive score, and select the optimal class selection scheme. S5. Combining loop control and scheme rollback strategy, the optimal shift removal scheme is iteratively processed to obtain the final shift removal scheme. S6. Verify and statistically analyze the final shift selection plan, and generate the corresponding shift selection report.

2. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, In S1, the shift allocation target includes the limit on the number of vehicles allocated per shift for the upbound journey obtained from the main station and the limit on the number of vehicles allocated per shift for the downbound journey obtained from the secondary station; The time period characteristics include the defined morning rush hour and evening rush hour; The scoring weights and penalty coefficients include: peak non-overlap weight α1, departure interval weight α2, mode feature weight β, vehicle number weight γ, passenger flow matching weight δ, operational stability weight ε, and first and last bus penalty coefficients; the mode feature weight β includes splitting and randomizing the shift weight mode β1 and last bus mode β2; the passenger flow matching weight includes density matching weight δ1, load factor weight δ2, and capacity balance weight δ3.

3. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, In S2, the decision variables are defined as follows: Define a vehicle set V = {v1, v2, ..., v...} n }, where n is the total number of vehicles; For vehicle v i Define its class set m i For vehicle v i The total number of flights; Define the decision variable x for random shifts ij x ∈{0,1} ij =1 indicates that the first [value] is retained. Each class, x ij =0 indicates that the shift has been deleted; Define direction indicator variable d i ∈{0,1},d i =0 indicates the upward direction, d i =1 indicates the downward direction.

4. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, In S2, establishing the objective function maxF(x) includes: ； Where, f1(x) ij f2(x) ij f3(x) ij f4(x) ij f5(x) ij f6(x) ij ) represent the scoring functions for peak hours, departure intervals, mode characteristics, vehicle number, passenger flow matching, and operational stability, respectively, p(x ij ) represents the penalty function, and w1, w2, w3, w4, w5, and w6 represent the weight coefficients for each dimension.

5. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, In S2, the constraints include: The continuity constraint is used to limit the continuity of the remaining shifts after shift selection. It requires that the number of consecutive segments does not exceed 2 and that the parity rule is satisfied. Directional constraints require that the direction of the earliest departure event of a vehicle after the shift is drawn must be consistent with the initial direction of the vehicle. Quantity constraints limit the total number of shifts drawn in both the upward and downward directions to a preset limit. The integrity constraint requires that the number of shifts retained for each vehicle after shift selection is no less than the minimum required number of shifts.

6. The bus shift selection optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, The multi-dimensional scoring system in S3 includes: Peak hour scoring is used to assess the impact of shift scheduling on peak hour operations; Departure interval score is used to evaluate the uniformity and rationality of departure intervals after random shifts are taken. Pattern feature scoring is used to identify and differentiate patterns such as split shifts, consecutive shifts, and those involving the last bus. Vehicle number scoring is used to assign specific bonus points based on the vehicle number to aid in decision-making. Passenger flow matching score is used to evaluate the degree of matching between the shift scheduling plan and passenger flow demand, including passenger flow density, load factor and capacity balance; Operational stability score is used to assess the impact of shift changes on vehicle utilization and operational stability; The first and last bus penalty score is used to impose severe penalties on schemes that remove the first or last bus.

7. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, In S3, the constraint verification strategy includes: Continuity check: Verify whether the remaining shift index forms a valid continuous segment and meets the parity requirement; Directional verification: Verify whether the direction of the earliest departure event after the shift selection is consistent with the initial direction of the vehicle. Completeness check to verify whether the plan after shift sampling meets the minimum number of shifts requirement and the rationality of the operation trajectory.

8. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, S4 includes: S41. Traverse the vehicle set, filter the vehicles to be evaluated based on directional constraints, calculate the target number of retained shifts based on the original number of shifts of the vehicles, generate candidate shift deletion combinations using a combination algorithm, and apply constraint verification strategy to filter out feasible candidate solutions. S42. For each feasible candidate solution, call the various scoring functions in the multi-dimensional scoring system to calculate its score, and combine them with preset weights to perform weighted summation to obtain the comprehensive score of the solution. S43. Sort all feasible candidate solutions in descending order of comprehensive score, and select the solution ranked first as the current optimal solution for class selection. S44. Based on the selected optimal solution, delete the specified shift from the corresponding vehicle shift table, update the vehicle status to single shift, and update the single shift vehicle allocation counter for the corresponding direction.

9. The bus shift optimization method based on multi-dimensional scoring as described in claim 1, characterized in that, S5 includes: S51. At the beginning of each iteration, determine whether the current total number of shifts drawn has reached the preset shift drawing target, and determine whether the number of shifts drawn in the upward and downward directions has reached their preset limits respectively; if any condition is met, terminate the iteration process. S52. When the current round of shift selection fails because a feasible solution cannot be generated or an invalid solution is generated, the vehicle shift schedule and related counters will be rolled back to the state before the current iteration. S53. After generating the candidate optimal shift selection scheme in each iteration, verify whether the scheme satisfies all mathematical constraints and business rules, and confirm the correctness of the vehicle status update. S54. For the verified scheme, update the system status, including the total number of shifts, the shift counters for each direction, and the vehicle status.

10. The method for optimizing bus shift selection based on multi-dimensional scoring as described in claim 1, characterized in that, In S6, the shift report includes: A summary of the shift change plan, listing all vehicles whose shifts were changed and the specific shifts that were deleted; The scoring analysis report shows the final solution's scoring details across various dimensions; A constraint satisfaction report, describing the satisfaction status of all mathematical and business constraints; Optimize the effect evaluation and provide comparative analysis of key operational indicators before and after shift changes.

Images