A substation inspection scheme planning method for extreme load scenarios

Abstract

The application relates to a substation inspection scheme planning method for an extreme load scene, which comprises the following steps: constructing a substation inspection model fusing a value-keeping scene; proposing a dynamic updating mechanism of an inspection path adapting to the change of the number of stations; establishing a time-cost collaborative planning model considering resource constraints to generate an optimal inspection scheme considering the inspection time length and fuel cost; defining work length, station inspection proportion, overtime intensity, work frequency and fatigue degree as work quantity evaluation indexes, constructing an inspection task allocation model, and optimizing and solving the substation inspection scheme with the most balanced work quantity allocation. The application constructs a multi-dimensional work quantity evaluation index system and an inspection task allocation model, realizes the balanced allocation of work quantities among various inspection teams, and provides a scientific decision basis for personnel scheduling in operation and maintenance management.

Description

Technical Field

[0001] This application relates to the field of substation inspection technology, specifically to a substation inspection scheme planning method for extreme load scenarios. Background Technology

[0002] With the rapid development of the economy and society, the electricity demand for industrial production and residential use continues to rise, and the peak load of the power grid is repeatedly breaking records. Especially during peak summer and winter seasons, key equipment such as main transformers in substations face the test of long-term, high-load operation, significantly increasing the risk of equipment overload and the probability of failure. How to efficiently, economically, and rationally plan cross-site inspection tasks under extreme load scenarios, and coordinate and allocate limited operation and maintenance human resources, has become a practical problem that urgently needs to be solved in current substation operation and maintenance work.

[0003] Currently, research on substation inspection optimization has made some progress both domestically and internationally, but it mainly focuses on inspection path planning at the substation level. This involves planning the shortest detection paths between equipment within a single substation for inspection robots, drones, or personnel, thereby improving the efficiency of single-station inspections. However, research on "inter-substation" inspection path planning covering multiple substations is limited. Especially under extreme grid load scenarios, inspection tasks often need to be closely integrated with on-site personnel monitoring requirements, but existing methods generally do not consider the monitoring scenario as a constraint, resulting in poor adaptability in practical applications.

[0004] Existing research on inspection route optimization typically aims to minimize total inspection time or path distance. However, in practice, both inspection efficiency (time cost) and economy (fuel cost) are crucial dimensions for evaluating a solution. Especially in inspection scenarios with wide geographical coverage and dispersed stations, fuel cost cannot be ignored. How to achieve a dynamic balance between inspection time and fuel cost has not yet been fully explored.

[0005] Regarding the adaptability of inspection paths, existing methods typically perform one-time optimization based on a static set of inspection sites, lacking the ability to dynamically update. In actual power grid operation, changes in equipment operating status or adjustments to power grid operation modes can lead to increases or decreases in the number of inspection sites. Existing methods do not fully account for this real-world condition and lack the ability to dynamically update and plan inspection paths in real time, making it difficult to adapt to complex and ever-changing inspection needs.

[0006] Furthermore, at the execution level of inspection tasks, existing methods have not yet established a scientific and comprehensive task allocation mechanism. In reality, task allocation often relies on team priority (such as backup one, backup two, backup three, etc.) or simple time averaging principles, which easily leads to uneven workloads, with some teams overloaded while others are underloaded, thus causing internal conflicts and dampening employee enthusiasm and team cohesion. Therefore, it is urgent to construct a scientific workload evaluation index system to achieve a balanced distribution of inspection tasks and provide a basis for personnel scheduling in operation and maintenance management. Summary of the Invention

[0007] To address the problems of insufficient adaptability of existing substation inspection methods under extreme load scenarios, difficulty in balancing economy and timeliness, static path updates, and uneven task allocation, this invention proposes a substation inspection scheme planning method for extreme load scenarios.

[0008] To achieve the above objectives, this application provides the following technical solution:

[0009] This application provides a method for planning substation inspection schemes for extreme load scenarios, including the following steps:

[0010] Step 1: Construct a substation inspection model that integrates on-duty scenarios, including a "single team travel" model that considers single-station on-duty work and a "multi-team travel" model that considers both single-station and dual-station on-duty work.

[0011] Step 2: Propose a dynamic update mechanism for inspection routes that adapts to changes in the number of stations, and formulate route update rules for the "single team travel" model and the "multiple teams traveling together" model respectively;

[0012] Step 3: Establish a time-cost collaborative planning model that considers resource constraints. By improving the density peak clustering algorithm, the patrol team's stationing points are selected and the patrol stations are clustered in a secondary manner. Then, differentiated optimization is performed on each type of cluster to generate the best patrol plan that takes into account both patrol time and fuel cost.

[0013] Step 4: Define workload evaluation indicators such as working hours, patrol ratio, overtime intensity, number of shifts and fatigue level, construct a patrol task allocation model, and optimize the substation patrol scheme with the most balanced workload allocation.

[0014] The “single-team travel” model is shown in equations (1) and (2):

[0015] (1)

[0016] In the formula: The time function of the "single team travel" model after considering the on-duty working conditions; Number the route; For the "single team travel" model under on-duty conditions, the route The total number of substations in the area; Number the substations on route r; For substation Inspection duration; This is a type of inspection, corresponding to a special inspection. For vehicles from the substation To the substation Average time required; For vehicles from the substation Time to return to the origin; These are auxiliary variables introduced for calculation and have no actual physical meaning.

[0017] (2)

[0018] In the formula: The cost function for the "single-team travel" model in the scenario p under duty monitoring; The duty scenario is numbered, with values ​​of 0, 1, and 2, representing no duty, single-station duty, and dual-station duty conditions, respectively. For the "single team travel" model, ; For duty scenarios Below is the total number of substations along route r in the "single-team travel" model; For substation With substation The distance between them; For substation Distance to the origin; This refers to the fuel consumption per 100 kilometers for engineering vehicles. For fuel prices.

[0019] The time functions of the "multi-team travel" model are shown in equations (3) and (5), respectively, and the cost function is shown in equation (7):

[0020] (3)

[0021] In the formula: To consider the time function of the "multi-team simultaneous" model after single-station duty; The total number of substations on route r in the "multi-team operation" model under single-station duty conditions; , , , , , Vehicles from the station arrive , arrive , arrive , arrive , Time required to reach the origin and time required to reach substation 2 from the origin; , Sites , The time compensation amount is calculated using formula (4).

[0022] (4)

[0023] In the formula: For the first time the vehicle leaves the substation The time elapsed from departure to the second arrival at the station is its value. .

[0024] (5)

[0025] In the formula: To consider the time function of the "multi-team simultaneous" model after dual-station duty; The total number of substations on route r in the "multi-team operation" model under dual-station duty conditions; , , Vehicles from the station arrive , arrive , The time required to return to the origin under dual-station monitoring conditions. The calculation method is shown in equation (6).

[0026] (6)

[0027] (7)

[0028] In the formula: , , These are the cost functions for the "multi-team simultaneous" model under the conditions of not considering on-duty personnel, single-station on-duty personnel, and dual-station on-duty personnel, respectively. The total number of substations on route r in the "multi-team parallel" model under unattended working conditions; , , , , , , , , , , , Sites arrive , arrive , arrive , arrive , arrive , arrive , arrive , arrive , arrive , To the origin 2. Distance from the origin to the origin.

[0029] The path update mechanism of the "single team travel" model is as follows: assuming the patrol team is inspecting the [number]th [unit / item], If a new site is added at this time Several stations await inspection. If a station is to be monitored, the time function of the original model will be updated as follows:

[0030] (8)

[0031] In the formula: The time function for the "single-team travel" model without considering on-duty situations; , , These represent the additional inspection time added to the original model under different working conditions; , , Vehicles from the station arrive , To the origin The time required to reach the origin; , Sites , Inspection duration.

[0032] Accordingly, the cost function of the "single-team travel" model is updated simultaneously:

[0033] (9)

[0034] In the formula: , These are the cost functions for the "single team travel" model under the scenarios of no staffing and single-station staffing, respectively. , Different working conditions Additional fuel costs; , , Sites arrive , To the origin The distance to the origin.

[0035] The path update mechanism of the "multi-team travel" model is as follows:

[0036] For the "multi-team simultaneous" model, the update principles of the model's time and cost functions are closely related to the types of newly added stations. When a patrol team conducts its first patrol... During the inspection of each station, if a new one is added One inspection station to be inspected For each station to be monitored, the time and cost functions of the model are updated according to equations (10) and (11), respectively:

[0037] (10)

[0038] In the formula: The time function for the "multi-team simultaneous" model under unattended conditions; , , These represent the additional inspection time added to the original model under different working conditions; , , , Vehicles from the station arrive , arrive , arrive , The time required to reach the origin; , Sites , The amount of time compensation.

[0039] (11)

[0040] In the formula: , , Different working conditions Additional fuel costs; , , , Sites arrive , arrive , arrive , The distance to the origin.

[0041] In response to changes in the number of stations, the dynamic update principle for inspection routes is as follows: The starting point is set according to the real-time location of the patrol team. If the patrol team is performing an inspection task within the station, the current station is used as the starting point; if the team is on the way to the next station, the next target station is used as the starting point. The origin is used as the endpoint, and the improved gold mining algorithm is used to re-plan the inspection routes for the remaining stations, updating the original inspection routes to the optimized new routes.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] A multi-dimensional workload evaluation index system and an inspection task allocation model were constructed, which realized the balanced allocation of workload among various inspection teams and provided a scientific decision-making basis for personnel scheduling in operation and maintenance management. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A flowchart for planning substation inspection schemes.

[0046] Figure 2 A schematic diagram of a "single-team travel" model for integrating duty scenarios.

[0047] Figure 3 A schematic diagram of a "multi-team collaboration" model for integrating duty scenarios.

[0048] Figure 4 This is a schematic diagram of the inspection path update based on the "single team travel" model.

[0049] Figure 5 This is a geographical distribution map of the substations under the jurisdiction of the centralized control station.

[0050] Figure 6 The results of secondary clustering of the sites to be inspected.

[0051] Figure 7 The variance of the workload evaluation index for each allocation strategy under different inspection models. Detailed Implementation

[0052] The technical solutions of the embodiments of this application will now be described with reference to the accompanying drawings. It should be noted that similar reference numerals and letters in the following drawings indicate similar items; therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0053] The terms “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0054] The terms “first,” “second,” etc., are used only to distinguish one entity or operation from another, and should not be construed as indicating or implying relative importance, nor as requiring or implying any such actual relationship or order between these entities or operations.

[0055] like Figure 1 As shown in the figure, this application provides a substation inspection scheme planning method for extreme load scenarios, characterized by the following steps:

[0056] Step 1: Construct a substation inspection model that integrates on-duty scenarios, including a "single team travel" model that considers single-station on-duty work and a "multi-team travel" model that considers both single-station and dual-station on-duty work.

[0057] Step 2: Propose a dynamic update mechanism for inspection routes that adapts to changes in the number of stations, and formulate route update rules for the "single team travel" model and the "multiple teams traveling together" model respectively.

[0058] Step 3: Establish a time-cost collaborative planning model that considers resource constraints. By improving the density peak clustering algorithm, the patrol team's stationing points are selected and the patrol stations are further clustered. Then, differentiated optimization is performed on each type of cluster to generate the best patrol plan that takes into account both patrol time and fuel cost.

[0059] Step 4: Define workload evaluation indicators such as working hours, patrol ratio, overtime intensity, number of shifts and fatigue level, construct a patrol task allocation model, and optimize the substation patrol scheme with the most balanced workload allocation.

[0060] "Single Team Travel" Inspection Model that Integrates with Duty Scenarios

[0061] The "single-team patrol" model (hereinafter referred to as the "single-team patrol" model) refers to a patrol method in which a group of patrol teams undertakes all station patrol tasks. It is suitable for areas where stations are scattered and task priorities differ significantly. By concentrating resources to complete high-priority tasks, it reduces manpower redundancy.

[0062] Since the traditional "single-team travel" model does not take into account the situation of station duty, the engineering vehicle, after considering the situation of station duty, will transport the patrol team to the duty station and then return on its own, no longer waiting for the patrol team to finish its work. The schematic diagram of the model is as follows. Figure 2 As shown, the mathematical model is as shown in equations (1) and (2):

[0063] (1)

[0064] In the formula: The time function of the "single team travel" model after considering the on-duty working conditions; Number the route; For the "single team travel" model under on-duty conditions, the route The total number of substations in the area; Number the substations on route r; For substation Inspection duration; For the inspection type, l is set to 1 in this paper, corresponding to a special inspection; For vehicles from the substation To the substation Average time required; For vehicles from the substation Time to return to the origin; These are auxiliary variables introduced for calculation and have no actual physical meaning.

[0065] (2)

[0066] In the formula: The cost function for the "single-team travel" model in the scenario p under duty monitoring; The duty scenario is numbered, with values ​​of 0, 1, and 2, representing no duty, single-station duty, and dual-station duty conditions, respectively. For the "single team travel" model, ; For duty scenarios Below is the total number of substations along route r in the "single-team travel" model; For substation With substation The distance between them; For substation Distance to the origin; This refers to the fuel consumption per 100 kilometers for engineering vehicles. For fuel prices.

[0067] "Multi-team simultaneous" inspection model integrating duty scenarios

[0068] The "multi-team coordinated" inspection model (hereinafter referred to as the "multi-team coordinated model"), which uses engineering vehicles to coordinate the transport of two patrol teams to complete substation inspection tasks, is suitable for areas with dense substations and high task overlap. It effectively reduces vehicle idle time and significantly saves time costs. The model's monitoring scenarios can be divided into single-station monitoring and dual-station monitoring. Taking dual-station monitoring as an example... Figure 2 A schematic diagram is provided.

[0069] Figure 3 During the exercise, each patrol team is transported to the duty station by an engineering vehicle. The specific procedure is as follows: after patrol team 1 completes the inspection task at station B, it is picked up by a vehicle and transported to station D for duty; then the vehicle goes to station C to transport patrol team 2 to station E for duty; after patrol team 2 arrives at station E, the vehicle will return to the original point on its own.

[0070] Taking into account the on-duty status, the time functions of the "multi-team travel" model are shown in equations (3) and (5), and the cost function is shown in equation (7):

[0071] (3)

[0072] In the formula: To consider the time function of the "multi-team simultaneous" model after single-station duty; The total number of substations on route r in the "multi-team operation" model under single-station duty conditions; , , , , , Vehicles from the station arrive , arrive , arrive , arrive , Time required to reach the origin and time required to reach substation 2 from the origin; , Sites , The time compensation amount is calculated using formula (4).

[0073] (4)

[0074] In the formula: For the first time the vehicle leaves the substation The time elapsed from departure to the second arrival at the station is its value. .

[0075] (5)

[0076] In the formula: To consider the time function of the "multi-team simultaneous" model after dual-station duty; The total number of substations on route r in the "multi-team operation" model under dual-station duty conditions; , , Vehicles from the station arrive , arrive , The time required to return to the origin. Under dual-station monitoring conditions, The calculation method is shown in equation (6).

[0077] (6)

[0078] (7)

[0079] In the formula: , , These are the cost functions for the "multi-team simultaneous" model under the conditions of not considering on-duty personnel, single-station on-duty personnel, and dual-station on-duty personnel, respectively. The total number of substations on route r in the "multi-team parallel" model under unattended working conditions; , , , , , , , , , , , Sites arrive , arrive , arrive , arrive , arrive , arrive , arrive , arrive , arrive , To the origin 2. Distance from the origin to the origin.

[0080] Dynamic update mechanism for inspection paths to adapt to changes in the number of stations

[0081] In practical engineering, the load rate of power equipment fluctuates in real time with factors such as time and temperature, and the number of substations to be inspected changes dynamically under extreme load scenarios. If the patrol team is on the field performing inspection tasks and the number of inspection sites increases or decreases, following the predetermined inspection route and then replanning a new inspection route after the inspection is completed will result in a significant waste of manpower and vehicle fuel costs, and a substantial increase in inspection time. Therefore, this paper constructs a dynamic update mechanism for inspection paths that adapts to changes in the number of inspection sites, taking the increase in the number of sites as an example, for different inspection models. The analysis method for the decrease in the number of sites is similar and will not be repeated due to space limitations.

[0082] "Single Team Travel" Model Path Update Mechanism

[0083] Figure 4 Based on the "single team travel" model, the changes in the patrol team's inspection route are shown when the number of stations increases. Figure 4 In the process, the patrol team's original patrol route was O→A→B→C→D→O. When patrolling to station B, station F was added to be patrolled. At this time, after real-time optimization by the algorithm, the patrol route was updated to O→A→B→E→F→G→O.

[0084] Suppose the patrol team is inspecting the first... If a new site is added at this time Several stations await inspection. If a station is to be monitored, the time function of the original model will be updated as follows:

[0085] (8)

[0086] In the formula: The time function for the "single-team travel" model without considering on-duty situations; , , These represent the additional inspection time added to the original model under different working conditions; , , Vehicles from the station arrive , To the origin The time required to reach the origin; , Sites , Inspection duration.

[0087] Accordingly, the cost function of the "single-team travel" model is updated simultaneously:

[0088] (9)

[0089] In the formula: , These are the cost functions for the "single team travel" model under the scenarios of no staffing and single-station staffing, respectively. , Different working conditions Additional fuel costs; , , Sites arrive , To the origin The distance to the origin.

[0090] "Multi-team travel" model path update mechanism

[0091] For the "multi-team simultaneous" model, the update principles of the model's time and cost functions are closely related to the type of newly added stations (attended or unattended). When a patrol team conducts its first patrol... During the inspection of each station, if a new one is added One inspection station to be inspected For each station to be monitored, the time and cost functions of the model are updated according to equations (10) and (11), respectively:

[0092] (10)

[0093] In the formula: The time function for the "multi-team simultaneous" model under unattended conditions; , , These represent the additional inspection time added to the original model under different working conditions; , , , Vehicles from the station arrive , arrive , arrive , The time required to reach the origin; , Sites , The amount of time compensation.

[0094] (11)

[0095] In the formula: , , Different working conditions Additional fuel costs; , , , Sites arrive , arrive , arrive , The distance to the origin.

[0096] In response to changes in the number of stations, the dynamic update principle for inspection routes is as follows: Starting points are categorized based on the real-time location of the patrol team—if the patrol team is performing an inspection task within a station, the current station is the starting point; if the team is en route for inspection, the next target station is the starting point; the origin is uniformly used as the endpoint, and the improved Gold Rush Optimizer (IGRO) algorithm is employed to replan the inspection routes for the remaining stations, updating the original inspection routes to the optimized new routes.

[0097] Time-cost collaborative programming model considering resource constraints

[0098] Site selection for outposts

[0099] To improve the power supply reliability and emergency response capabilities of the power grid under extreme load scenarios, control stations or substation work areas typically rely on site resources to coordinate and dispatch patrol teams to stations near load centers with convenient transportation for rapid response and efficient handling of emergencies. Considering both engineering realities and objective conditions, this paper proposes an improved Density Peaks Clustering (DPC) algorithm for selecting stationing points. The improvements are mainly reflected in the following aspects:

[0100] First, specify the number of cluster centers. The central control station or substation area is designated as one of the cluster centers by default. Select cluster centers :

[0101] (12)

[0102] In the formula: Represents a centralized control station or substation area; The set of sites sorted in descending order of local density; Let be the local density of the (k-1)th station; This is the (k-1)th station; The number of cluster centers optimized for the DPC algorithm; This is the union symbol.

[0103] Secondly, based on the characteristics of the substation's geographical coordinate data, a preset quantile threshold is used, and the truncation parameter is calculated with reference to formula (13) to avoid the problems of density distribution sparsity and discrimination imbalance caused by subjective selection of parameters.

[0104] (13)

[0105] In the formula: The set of non-repeating distances extracted; This is the geographic coordinate matrix of the substation; The horizontal axis number; The vertical axis represents the index; N represents the total number of substations. It is a quantile function; This is the quantile threshold; This is for truncating parameters.

[0106] Planning Model

[0107] objective function

[0108] Centered on the selected garrison points, the stations are clustered a second time based on the local density of the stations to be inspected, and the density of each type of cluster is calculated according to equation (18). If the density of a certain type of cluster is less than If the cluster is in a certain condition, the "multi-team travel" model is used for planning; otherwise, the "single-team travel" model is used.

[0109] (14)

[0110] In the formula: Cluster density; Assign cluster numbers; This represents the total number of stations to be inspected. Stations awaiting inspection and The distance between them; , All are station numbers.

[0111] For different inspection models, the two indicators of inspection time and fuel cost are weighted differently. The objective function is constructed as shown in Equation (15) with the shortest time and the lowest cost as the optimization objectives, and IGRO is used to solve it.

[0112] (15)

[0113] In the formula: The overall objective function; This is a function to find the maximum value. The total number of clusters; , Let e ​​be two sub-objective functions for cluster e, representing inspection time and fuel cost respectively, where inspection time is determined by the vehicle with the longest travel time; Let e ​​be the set of vehicles belonging to cluster e; for Vehicles in the middle; For vehicles Travel duration; They are respectively and The set of elements that make up the composition, where This represents the number of algorithm iterations. , These are the weights for inspection time and fuel cost indicators, respectively, to meet the requirements. ; This represents the total number of routes corresponding to the guarded scenario p in cluster e; The cost function for the "multi-team traveling together" model in the scenario p under the duty system; For the "multiple teams traveling together" model route r under the scenario p, the total number of substations.

[0114] Constraints

[0115] 1) Constraints on the number of guard stations

[0116] (16)

[0117] In the formula: The total number of patrol teams in cluster e; The total number of guard stations in cluster e; The number of stations on route r of cluster e using the "single-team travel" model; This represents the number of stations on route r of cluster e, which adopts the "multi-team travel" model.

[0118] 2) Route start and end point constraints

[0119] A route is defined as a single closed loop path through which vehicles travel from the starting point to the ending point, and the starting and ending points of the route are specified as the origin, i.e.:

[0120] (17)

[0121] In the formula: , Both are binary decision variables; if the vehicle From the origin to station g, then Otherwise, it is 0. If the vehicle... From the site To the origin, then Otherwise, it is 0; g, The number of the inspection station on a certain route; This represents the total number of vehicles in cluster e.

[0122] 3) Route travel time constraints

[0123] For different inspection models, preset inspection route travel time limits are required. Each model should satisfy:

[0124] (18)

[0125] In the formula: , These are the time functions of route r in the "single team travel" and "multiple teams traveling together" models under the duty scenario p.

[0126] 4) Constraints on the inspection sequence of the duty station

[0127] For clusters using the "single-team travel" model, if a patrol route includes a guard station s, then that station is set as the final patrol station on the route, and the vehicle returns to the origin from that station. Simultaneously, the total number of patrol teams decreases by one.

[0128] (19)

[0129] In the formula: For binary decision variables, if vehicle From station s to the origin, then Otherwise, it is 0; This represents the total number of stations to be inspected in cluster e.

[0130] For clusters using the "multi-team traveling together" model: if a patrol route includes a guard station s, then that station is designated as the second-to-last station on the route. After completing the inspection of the next station, the vehicle returns to the origin, and the total number of patrol teams decreases by 1. The rule model is shown in equation (20); if a patrol route includes both guard stations s and If the two stations at the end of the route are set, the vehicles will return to the origin after arriving at the two stations in sequence. The total number of patrol teams will be reduced by 2 accordingly. The model is shown in equation (21).

[0131] (20)

[0132] (twenty one)

[0133] In the formula: , Both are binary decision variables; if the vehicle From the duty station To the origin, then Otherwise, it is 0. If the vehicle... From station g to the duty station ,but Otherwise, it is 0.

[0134] 5) Flow balance constraints

[0135] For any station g in cluster e, the number of times a vehicle enters should equal the number of times it exits; at the same time, the number of visits to unattended station h must meet the cluster type requirements, as shown in equation (22):

[0136] (twenty two)

[0137] In the formula: , , Both are binary decision variables; if the vehicle From the site To station g, then Otherwise, it is 0. If the vehicle... From site g to site ,but Otherwise, it is 0. If the vehicle... From site g to site h, then Otherwise, it is 0.

[0138] Inspection task allocation strategy and overall planning process considering workload balance

[0139] Based on the optimal inspection schemes for different clusters selected through optimization, and combined with the vehicle and personnel configurations for each cluster, workload evaluation indicators are constructed to optimize the allocation of inspection tasks.

[0140] Workload Evaluation Indicators

[0141] For cluster e, based on the patrol team set Inspection route assembly Route inspection duration set and the set of route station numbers Define travel decision variables It takes the value 0 or 1, when The team is responsible for inspecting the route. hour, Otherwise, it is 0. Symbol explanation: The elements 1, 2, 3, and 4 represent standby team 1, standby team 2, standby team 3, and the duty team, respectively. for The elements in the text represent a patrol team; , , These are the set of inspection routes, the set of route inspection durations, and the set of route station numbers for cluster e in the optimal inspection scheme, respectively. , , The following are the conditions for cluster e under unattended, single-station-attended, and dual-station-attended working conditions, respectively. , , One route; , , All are route numbers; , , The following are the conditions for cluster e under unattended, single-station-attended, and dual-station-attended working conditions, respectively. , , Inspection time for each route; , , The following are the conditions for cluster e under unattended, single-station-attended, and dual-station-attended working conditions, respectively. , , The number of stations on the route; In the scenario p, the first cluster e One route.

[0142] 1) Working hours

[0143] The working hours of the patrol team depend on two factors: the duration of the patrol route itself and whether it is necessary to stay on duty at the station. The calculation method is shown in equation (23):

[0144] (twenty three)

[0145] In the formula: For cluster e Group work hours metrics; In the scenario p, the first cluster e Inspection time for each route; As a valued decision variable, it takes the value 0 or 1. Group 1 in When on duty along this route, Otherwise, it is 0.

[0146] 2) Station patrol ratio

[0147] The proportion of the number of stations actually inspected by the patrol team to the total number of stations to be inspected in its respective cluster is defined as the patrol ratio, i.e.:

[0148] (twenty four)

[0149] In the formula: For cluster e The proportion of patrol stations in the group; For the scenario p under the supervision, the first cluster e The number of stations on the route; It is a rounding function; This is a random variable for the algorithm, taking the value 0 or 1.

[0150] 3) Overtime intensity

[0151] During the patrol team assembly, except for the duty group, the daily working hours for all other groups are stipulated to be 8 hours. Based on this, the amount of overtime is defined by formula (25), and then an overtime intensity index is constructed:

[0152] (25)

[0153] (26)

[0154] In the formula: For cluster e Group's work overtime; For cluster e The overtime intensity index for the group.

[0155] 4) Number of workdays

[0156] The number of work trips refers to the number of times each team goes out with the vehicle, which is consistent with the number of inspection routes undertaken. The calculation method is based on formula (27):

[0157] (27)

[0158] In the formula: For cluster e The number of times the group works.

[0159] 5) Fatigue level

[0160] The fatigue level of each team is comprehensively assessed using the above indicators. After normalization to eliminate differences in the dimensions of the indicators, weights are assigned according to the differences in the importance of each indicator, and a weighted sum is obtained to obtain the fatigue level index:

[0161] (28)

[0162] In the formula: For cluster e Group fatigue level indicators; , , , Indicators , , , The normalized value; , , , The weights for the indicators are set to 0.45, 0.2, 0.25, and 0.1 in this paper.

[0163] Inspection task allocation model

[0164] Based on the workload evaluation index, with the goal of balancing and rationally allocating the patrol tasks of the patrol teams, an objective function and constraints are constructed, and the optimal patrol scheme for each team is solved using IGRO.

[0165] objective function

[0166] This paper aims to minimize the sum of variances of all indicators, that is:

[0167] (29)

[0168] In the formula: Assign the objective function of the model to the task; This is the function for taking the average value.

[0169] Constraints

[0170] 1) Travel group constraints

[0171] For different clusters, the inspection route The number of patrol teams should meet the following requirements:

[0172] (30)

[0173] 2) Restrictions on the duty team

[0174] Define variables Characterization route The number of guard teams on duty, which can be 0, 1, or 2, is as follows:

[0175] (31)

[0176] After the vehicle has completed the route Afterwards, the total number of patrol teams The following should be updated:

[0177] (32)

[0178] 3) Team priority constraints

[0179] For guard posts, the selection of guard teams is based on the priority of each patrol team. (Regarding patrol teams...) Define the team priority relationship as shown in equation (33):

[0180] (33)

[0181] In the formula: Priority matrix for patrol teams; , , , These are the priority levels for backup teams (1, 2, 3) and the duty group in a monitored scenario, respectively, to meet the following requirements. .

[0182] Overall planning process

[0183] The process of planning a substation inspection scheme is as follows: Figure 1 As shown, the overall process is divided into four stages. The first stage involves site selection and clustering, where cluster centers are selected as inspection points through manual designation and cluster analysis, and secondary clustering of the substations to be inspected is performed based on these points. The second stage focuses on time-cost optimization, taking into account constraints such as vehicle, team, and site inspection sequence, and implementing differentiated optimization for various cluster models to select the substation comprehensive inspection plan with the shortest inspection time and fuel cost. The third stage involves real-time updating of the inspection plan, dynamically adjusting the inspection routes based on newly added site information. The fourth stage optimizes the allocation of inspection tasks. For different clusters, based on workload evaluation indicators and constraints such as team priority and the number of on-duty teams, the inspection tasks in the comprehensive plan are optimized and allocated to determine the final substation inspection plan with balanced workload.

[0184] Case Analysis

[0185] To verify the effectiveness and universality of the planning method presented in this paper, simulation analysis under different operating conditions was conducted using a centralized control station in a certain city as an example, based on the MATLAB simulation platform. Figure 5 A geographical distribution map of the substations under the jurisdiction of this central control station is provided. The following algorithm parameters, vehicle parameters, and inspection route travel time limits are preset: , Total number of gold prospectors: 50; Number of iterations: 200; , , The improved density peak clustering algorithm selected three guard posts: Station 1, Station 18, and Station 41. Each guard post was equipped with three engineering vehicles and four patrol teams: Backup 1, Backup 2, Backup 3, and the duty team.

[0186] Secondary clustering and differential optimization

[0187] Based on the three stationing points, the assumed inspection sites in Table 1 were clustered a second time. The clustering results are as follows: Figure 6 As shown. Cluster 1 is optimized using a "multi-team travel" model, with the weights of the inspection time indicator and the fuel cost indicator set to 0.4 and 0.6 respectively; clusters 2 and 3 are optimized using a "single-team travel" model. , Take values ​​of 0.6 and 0.4 respectively.

[0188] Table 1 Information on sites to be inspected

[0189]

[0190] Clustering results comparison

[0191] To demonstrate the advantages of secondary clustering of sites, differentiated optimizations were implemented based on the new cluster and the original cluster, and the optimization results were compared. The information of sites to be inspected in the original cluster is shown in Table 2, and the optimization results are shown in Table 3.

[0192] Table 2 Information on inspection sites in the original cluster

[0193]

[0194] Table 3 Optimization results of the original and new clusters

[0195]

[0196] As shown in Table 3, the objective function value of the optimized scheme based on the new cluster is smaller than that of the original cluster, making it a superior substation inspection scheme in terms of both inspection time and fuel cost. This advantage stems from the secondary clustering strategy. For the sites to be inspected, this strategy considers the geographical distribution characteristics of the sites in real time, reclassifies the clusters to which the sites belong, and flexibly adjusts travel strategies, breaking the inherent travel patterns of the original clusters and achieving a dual improvement in the economic efficiency and effectiveness of the inspection scheme.

[0197] Comparison of optimization strategies

[0198] The differentiated optimization strategy can fully leverage the travel advantages of each model while also mitigating its inherent weaknesses to some extent. To highlight the effectiveness of this strategy, based on the secondary clustering results, differentiated optimization and conventional optimization were implemented for its sub-clusters, and the optimization results were compared, as shown in Table 4. Specifically: Conventional optimization... ; Let be the objective function value for cluster e.

[0199] Table 4. Optimization results for various cluster types under different optimization strategies

[0200]

[0201] Table 4 shows that the objective function values ​​of all clusters based on the differentiated optimization strategy are smaller than those of the non-differentiated optimization strategy. This strategy combines the cost advantage of the "single-team travel" model with the time advantage of the "multi-team travel" model, reasonably adjusts the model weights, effectively weakens the model's shortcomings, and optimizes and selects an inspection scheme with better overall performance.

[0202] Inspection route dynamic update mechanism test

[0203] To test the performance of the mechanism under different inspection models, a scenario was set up: a new batch of inspection stations were added at 53 minutes, and the station information is shown in Table 5. Based on this scenario, the mechanism was compared with the following methods in simulation, and the comparison results are shown in Table 6. Method 1: Proximity Allocation Method. The distance between the new station and all uninspected stations was calculated sequentially, and the new station was assigned to the inspection route of the nearest uninspected station. After the patrol team completed the inspection task of the neighboring station, the inspection work of the new station was carried out immediately. Method 2: Successive Allocation Method. The route was replanned for the new stations, and the patrol teams other than the team with the longest inspection time and the duty team were assigned to the inspection tasks of these stations. After each team completed its original inspection task, the inspection work was carried out according to the newly planned route.

[0204] Table 5 Information on Newly Added Sites

[0205]

[0206] Table 6 Inspection route planning results under different methods

[0207]

[0208] Table 6 shows the inspection paths of the four patrol teams in different clusters. At the 53rd minute, the inspection trajectories of the four patrol teams in cluster 1 were 18→17→18, 18→23, 18→21, and 18→21→22, with corresponding new path starting points of 18, 23, 21, and 22, respectively; the inspection trajectories of the four patrol teams in cluster 3 were 41→43, 41→, 41→, and 41→, with corresponding new path starting points of 43, 52, 35, and 51, respectively. "→" indicates that the team is in motion. The results in Table 6 verify the superiority of the proposed method: it can achieve dynamic and global planning of inspection routes based on the real-time location of the patrol teams. Compared with Method 1 and Method 2, it has better performance and produces more ideal planning results. It overcomes the limitations of subjective experience-based decision-making and eliminates the unnecessary time consumption caused by continuous inspections, providing a more efficient solution for updating inspection paths.

[0209] Optimization of inspection task allocation

[0210] In practical engineering, task allocation typically employs two principles: The first is priority-oriented allocation. Tasks are sorted from longest to shortest duration, allocated sequentially by backup team 1, backup team 2, backup team 3, and shift team, with higher-priority teams undertaking longer tasks. The second is average task allocation. Tasks are allocated equally among teams based on total task duration, without considering team priority. To fully verify the effectiveness of the proposed task allocation strategy, inspection task allocation tests were conducted based on the optimization results of the new cluster, estimating the workload evaluation indicators for each team under different allocation strategies for different inspection models. The route information for different inspection models in the new cluster is shown in Table 7, and the corresponding evaluation results are shown in Tables 8 and 9. Figure 7 As shown.

[0211] Table 7 Inspection route information for different inspection models

[0212]

[0213] Table 8 Evaluation results of each model under different allocation strategies

[0214]

[0215] Table 8 shows that the objective function value of the patrol route allocation scheme optimized based on the strategy in this paper is lower than that of Strategy 1 and Strategy 2, indicating that the scheme achieves a balanced distribution of workload among patrol teams. Figure 7 The paper further demonstrates the variance of workload evaluation indicators for the three allocation strategies under different inspection models. It can be seen that the variance of each indicator for the strategy proposed in this application is significantly lower than that of the other two strategies. Its core advantage lies in using workload indicators as the allocation standard and relying on algorithmic optimization scheduling to achieve a balanced distribution of inspection tasks among the teams. This provides a reliable theoretical solution for resolving team conflicts caused by uneven task allocation in actual operation and maintenance work.

[0216] 1) The proposed "single team travel" and "multiple team travel" inspection models effectively integrate personnel duty scenarios, can flexibly adapt to single or double station duty requirements, and improve the adaptability and practicality of inspection solutions.

[0217] 2) By improving the DPC algorithm, secondary clustering of inspection stations was completed, and a differentiated optimization strategy was adopted for each type of cluster. Simulation results show that secondary clustering can take into account the geographical distribution characteristics of the stations in real time, dynamically adjust the inspection model, and break the inherent limitations of the original cluster division. Differentiated optimization can effectively balance inspection time and fuel cost, and obtain an inspection scheme with better overall performance.

[0218] 3) In response to changes in the number of inspection stations, the proposed dynamic update mechanism for inspection routes can flexibly adjust the inspection routes based on the current location of the patrol team, reducing unnecessary time consumption and resource waste, and improving response flexibility and overall inspection efficiency.

[0219] 4) A multi-dimensional workload evaluation index system and inspection task allocation model were constructed, realizing a balanced allocation of workload among various patrol teams, and providing a scientific decision-making basis for personnel scheduling in operation and maintenance management.

[0220] The planning method proposed in this application can cover most daily inspection scenarios. In the future, we can consider integrating multi-source dynamic information such as real-time weather and traffic conditions to build a collaborative model of manual inspection and "air-space-ground" integrated intelligent inspection, and explore in depth the collaborative mechanism and scheduling strategy among multiple inspection entities.

[0221] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for planning substation inspection schemes for extreme load scenarios, characterized in that, Includes the following steps: Step 1: Construct a substation inspection model that integrates on-duty scenarios, including a "single team travel" model that considers single-station on-duty work and a "multi-team travel" model that considers both single-station and dual-station on-duty work. Step 2: Propose a dynamic update mechanism for inspection routes that adapts to changes in the number of stations, and formulate route update rules for the "single team travel" model and the "multiple teams traveling together" model respectively; Step 3: Establish a time-cost collaborative planning model that considers resource constraints. By improving the density peak clustering algorithm, the patrol team's stationing points are selected and the patrol stations are clustered in a secondary manner. Then, differentiated optimization is performed on each type of cluster to generate the best patrol plan that takes into account both patrol time and fuel cost. Step 4: Define workload evaluation indicators such as working hours, patrol ratio, overtime intensity, number of shifts and fatigue level, construct a patrol task allocation model, and optimize the substation patrol scheme with the most balanced workload allocation.

2. The substation inspection scheme planning method for extreme load scenarios according to claim 1, characterized in that, The "single-team travel" model is shown in equations (1) and (2): (1) In the formula: The time function of the "single team travel" model after considering the on-duty working conditions; Number the route; For the "single team travel" model under on-duty conditions, the route The total number of substations in the area; Number the substations on route r; For substation Inspection duration; This is a type of inspection, corresponding to a special inspection. For vehicles from the substation To the substation Average time required; For vehicles from the substation Time to return to the origin; These are auxiliary variables introduced for calculation and have no actual physical meaning. (2) In the formula: The cost function for the "single-team travel" model in the scenario p under the duty system; The duty scenario is numbered, with values ​​of 0, 1, and 2, representing no duty, single-station duty, and dual-station duty conditions, respectively. For the "single team travel" model, ; For duty scenarios The following is the total number of substations on route r in the "single-team travel" model; For substation With substation The distance between them; For substation Distance to the origin; This refers to the fuel consumption per 100 kilometers for engineering vehicles. For fuel prices.

3. The substation inspection scheme planning method for extreme load scenarios according to claim 1, characterized in that, The time functions of the "multi-team traveling together" model are shown in equations (3) and (5), respectively, and the cost function is shown in equation (7): (3) In the formula: To consider the time function of the "multi-team simultaneous" model after single-station duty; The total number of substations on route r in the "multi-team operation" model under single-station duty conditions; , , , , , Vehicles from the station arrive , arrive , arrive , arrive , Time required to reach the origin and time required to reach substation 2 from the origin; , Sites , The time compensation amount is calculated using formula (4). (4) In the formula: For the first time the vehicle leaves the substation The time elapsed from departure to the second arrival at the station is its value. . (5) In the formula: To consider the time function of the "multi-team simultaneous" model after dual-station duty; The total number of substations on route r in the "multi-team operation" model under dual-station duty conditions; , , Vehicles from the station arrive , arrive , The time required to return to the origin under dual-station monitoring conditions. The calculation method is shown in equation (6). (6) (7) In the formula: , , These are the cost functions for the "multi-team simultaneous" model under the conditions of no on-duty personnel, single-station on-duty personnel, and dual-station on-duty personnel, respectively. The total number of substations on route r under the "multi-team parallel" model in the case of unattended operation. , , , , , , , , , , , Sites arrive , arrive , arrive , arrive , arrive , arrive , arrive , arrive , arrive , To the origin 2. Distance from the origin to the origin.

4. The substation inspection scheme planning method for extreme load scenarios according to claim 1, characterized in that, The path update mechanism of the "single team travel" model is as follows: assuming the patrol team is inspecting the [number]th [unit / item], [the path update mechanism is as follows]. If a new site is added at this time Several stations await inspection. If a station is to be monitored, the time function of the original model will be updated as follows: (8) In the formula: The time function for the "single-team travel" model without considering on-duty personnel; , , These represent the additional inspection time added to the original model under different working conditions; , , Vehicles from the station arrive , To the origin The time required to reach the origin; , Sites , Inspection duration. Correspondingly, the cost function of the "single-team travel" model is updated accordingly: (9) In the formula: , These are the cost functions for the "single team travel" model under the scenarios of no staffing and single-station staffing, respectively. , Different working conditions Additional fuel costs; , , Sites arrive , To the origin The distance to the origin.

5. The substation inspection scheme planning method for extreme load scenarios according to claim 1, characterized in that, The path update mechanism for the "multi-team traveling together" model is as follows: For the "multi-team traveling together" model, the update principles of the model's time and cost functions are closely related to the types of newly added stations. When a patrol team is conducting its first patrol, the update mechanism will update the path accordingly. During the inspection of each station, if a new one is added One inspection station to be inspected For each station to be monitored, the time and cost functions of the model are updated according to equations (10) and (11), respectively. (10) In the formula: The time function for the "multi-team simultaneous" model under unattended conditions; , , These represent the additional inspection time added to the original model under different working conditions; , , , , respectively, vehicles from the station arrive , arrive , arrive , The time required to reach the origin; , Sites , Time compensation amount, (11) In the formula: , , Different working conditions Additional fuel costs; , , , Sites arrive , arrive , arrive , Distance to the origin In response to changes in the number of stations, the dynamic update principle for inspection routes is as follows: The starting point is set according to the real-time location of the patrol team. If the patrol team is performing an inspection task within the station, the current station is used as the starting point; if the team is on the way to the next station, the next target station is used as the starting point. The origin is used as the endpoint, and the improved gold mining algorithm is used to re-plan the inspection routes for the remaining stations, updating the original inspection routes to the optimized new routes.

Images