A method, system and storage medium for dispatching electric power marketing work orders

Abstract

The present invention relates to the field of power marketing, and in particular to a method, system and storage medium for dispatching power marketing work orders. In the present invention, by setting the correspondence between work order categories and job types, maintenance personnel are divided into different job types according to their professional directions and levels, ensuring that each work order can be assigned to the corresponding maintenance personnel according to the category; the present invention combines the maintenance personnel's task progress, speed and remaining working hours to calculate the surplus threshold, so as to dispatch work orders according to the maintenance personnel's work saturation, and dispatch work orders with the shortest path as the goal. The present invention solves the defects of the prior art of long work order processing time and large efficiency gap, greatly improves the work order dispatch efficiency and actual processing efficiency, and reduces the processing time gap between different work orders.

Description

Technical Field

[0001] The present invention relates to the field of power marketing, and in particular to a method, system and storage medium for dispatching power marketing work orders. Background Art

[0002] With the continuous expansion of the power marketing business, the types of work orders have gradually increased, and demand has grown and changed rapidly. Power work orders are divided into multiple technical directions based on needs such as power supply and maintenance. Different technical directions are further divided into multiple levels due to the scope of power network equipment involved. Due to the wide variety of power work orders, traditional power marketing business work order processing is facing multiple problems such as fragmented work order business, overlapping work order processing channels, and slow response speeds. At the same time, some customer requests are handled separately, online requests cannot be managed in a unified closed-loop manner, and service standards are not unified. This leads to large differences in work order processing efficiency and long work order processing times, which seriously affects customer service satisfaction. Summary of the Invention

[0003] In order to solve the defects of long work order processing time and large efficiency gap in the above-mentioned existing technology, the present invention proposes a power marketing work order dispatching method, which greatly improves the work order dispatching efficiency and actual processing efficiency, and reduces the processing time gap between different work orders.

[0004] The present invention proposes a method for dispatching power marketing work orders, comprising the following steps:

[0005] S1. Set work order categories and job types, with a one-to-one correspondence between work order categories and job types. Job types include the professional field and level of maintenance personnel. Set up a work order classification engine, which is used to classify received work orders. Each work order category is associated with a corresponding expected time.

[0006] S2. Set the working period and execution path. The working period is a continuous time period. The execution path corresponds to the maintenance personnel one-to-one. The execution path is a traffic path starting from the maintenance personnel's current location and passing through the locations where the maintenance personnel reported unfinished work orders. Each work order on the execution path is associated with an expected completion time. Set the surplus weight a, a=b1×b2+c1×c2+d1×d2; b1 is the number of unfinished maintenance work orders, c1 is the actual completion time of the most recently completed work order minus the expected completion time, and d1 is the absolute value of the difference between the current time and the end time of the current execution path; b2, c2, and d2 are weights, and are all set values; the end time of the current execution path is the expected completion time of the last work order on the current execution path;

[0007] Work orders are assigned for each work order category. The work order distribution for a single work order category specifically includes the following steps S3-S4:

[0008] S3. At the starting point of each work period, select the n(k) work orders with the earliest reporting time from the unassigned work orders of the I(k)th work order as the starting work order, and assign the n(k) starting work orders to each maintenance personnel of the work type corresponding to the I(k)th work order, where n(k) is the total number of on-duty maintenance personnel of the work type corresponding to the I(k)th work order; 1≤k≤K; K is the total number of work order categories set manually, which is also the total number of work types; let the maintenance personnel's current location be the starting point and the report address of the maintenance personnel's corresponding starting work order be the end point to construct the maintenance personnel's execution path;

[0009] S4. For the undispatched work orders of category I(k), select the undispatched work orders as dispatch targets in descending order of reporting time. For each dispatch target, first determine whether there is a maintenance personnel who satisfies both the following conditions F1 and F2:

[0010] F1: The length of the shortest path between the reported location of the dispatched object and the execution path of the maintenance personnel is less than or equal to the set path threshold;

[0011] F2: Among the maintenance personnel who meet the condition F1, there are maintenance personnel whose surplus weight is greater than or equal to the set surplus threshold;

[0012] If there is a maintenance personnel who meets the above conditions F1 and F2, the dispatch object is inserted into the execution path of one of the maintenance personnel who meets the conditions;

[0013] If there is no maintenance personnel who meets the above conditions F1 and F2, the dispatch object will be assigned to a maintenance personnel of the corresponding work type and used as the end point of the maintenance personnel's execution path, or wait for the starting point of the next work period and assign the dispatch object as an undispatched work order;

[0014] Each work order on the execution path is associated with an expected completion time. The expected completion time of the starting work order is the start time of the corresponding work period plus the expected time corresponding to the starting work order; the expected completion time of the dispatch object inserted between any two work orders on the execution path is the expected completion time of the next work order of the dispatch object on the execution path minus half of the expected time corresponding to the dispatch object; the expected completion time of the dispatch object as the end point of the execution path is the expected completion time of the previous work order of the dispatch object plus the expected time corresponding to the dispatch object.

[0015] Preferably, in S4, if there are maintenance personnel who meet the above conditions F1 and F2, the maintenance personnel with the largest corresponding surplus weight a is selected from the maintenance personnel who meet the conditions F1 and F2 as the target personnel, and the dispatch object is inserted into the execution path of the target personnel.

[0016] Preferably, the method for inserting the dispatch object into the execution path of the target person is: calculating the path length between the reporting address of the dispatch object and the reporting addresses of each work order on the current execution path of the target person as a reference value, calculating the sum of the reference values ​​corresponding to the two adjacent work order reporting addresses on the current execution path of the target person, obtaining the two work orders corresponding to the minimum sum of the reference values ​​as reference objects, and inserting the dispatch object between the two reference objects to obtain a new execution path.

[0017] Preferably, in S4, if there are maintenance personnel who meet the above conditions F1 and F2, the maintenance personnel who meet the conditions F1 and F2 will be recorded as alternative personnel, and the execution path of the alternative personnel will be recorded as the alternative path. For each alternative path, the path length between the reporting address of the dispatch object and the reporting addresses of each work order on the alternative path is calculated as the reference value, the sum of the reference values ​​corresponding to the two adjacent work order reporting addresses on the alternative path is calculated, and the minimum sum of the reference values ​​is selected as the screening condition value for the dispatch object. The alternative path with the minimum screening condition value is selected as the target path, the dispatch object is assigned to the maintenance personnel corresponding to the target path, and the reporting address of the dispatch object is inserted between the two work order reporting addresses corresponding to the minimum sum of the reference values ​​on the target path to form a new execution path.

[0018] Preferably, in S4, if there is no maintenance personnel meeting the above conditions F1 and F2 for the dispatch object, the following sub-steps are performed:

[0019] S41. Obtain the end time corresponding to each current execution path. The end time is calculated as follows:

[0020] Ts(k)=T0+n'(k)×T(k);

[0021] Where Ts(k) is the end time of any execution path in the work type corresponding to the I(k)th work order, T0 is the current time, n'(k) is the number of work orders remaining on the execution path, and T(k) is the expected time associated with the I(k)th work order;

[0022] S42. Determine whether there is an execution path where Ts(k)+T(k) is less than or equal to the end point of the current work period; if not, wait until the start point of the next work period and assign the dispatch object as an undispatched work order;

[0023] If yes, then obtain all execution paths that satisfy Ts(k)+T(k) less than or equal to the end point of the current working period as alternative paths, calculate the path length between the end point of each alternative path and the reported address of the dispatch object, select the alternative path with the smallest corresponding path length as the target path, assign the dispatch object to the maintenance personnel corresponding to the target path, and update the end point of the target path to the reported address of the dispatch object.

[0024] Preferably, at the end of each work period, step S5 is executed: unfinished work orders on each execution path are collected, and at the starting point of the next work period, the collected unfinished work orders are distributed as undispatched work orders.

[0025] The present invention also provides a system and storage medium for carrying the above-mentioned power marketing work order dispatching method to facilitate the promotion of the method.

[0026] The present invention provides a power marketing work order dispatching system, comprising a memory and a processor, wherein the memory stores a computer program, the processor is connected to the memory, and the processor is used to execute the computer program to implement a power marketing work order dispatching method.

[0027] The present invention provides a storage medium storing a computer program, which is used to implement a method for dispatching power marketing work orders when executed.

[0028] The advantages of the present invention are:

[0029] (1) In the present invention, by setting up a correspondence between work order categories and types of work, maintenance personnel are divided into different types of work according to their professional direction and level. The professional direction indicates the technical direction in which the maintenance personnel are proficient, and the level indicates the maintenance personnel's authority, that is, the range of tools that the maintenance personnel can use and the operation authority of the power network components. In this way, the work order categories are divided according to the types of work, ensuring that each work order can be assigned to the corresponding maintenance personnel according to the category, so that it can be smoothly executed.

[0030] When allocating work orders, the present invention also incorporates electronic maps to plan execution paths, helping maintenance personnel save travel time and improving maintenance efficiency. Furthermore, the execution paths in the present invention are recommendations. When calculating maintenance personnel's work progress, the present invention incorporates the number of remaining work orders along the execution path for evaluation, avoiding the inflexibility of rigid execution paths and facilitating maintenance personnel to plan the order of work order processing based on objective factors such as their own habits. While ensuring the rational allocation of work orders, this provides maintenance personnel with a certain degree of autonomy, which helps to increase their subjective motivation, thereby further improving work order processing efficiency while ensuring the rational distribution of work orders.

[0031] (2) The surplus threshold provided by the present invention is calculated in combination with the maintenance personnel's task progress, speed and remaining working hours, which can accurately evaluate the saturation level of the maintenance personnel's current work order allocation, making it easier to subsequently dispatch work orders based on the maintenance personnel's work saturation, thereby ensuring that each work order can be processed in a timely manner.

[0032] (3) The present invention marks the expected completion time for each work order on the execution path. In particular, when a new work order is assigned, the expected completion time of the newly inserted work order is calculated based on the expected completion time of the remaining work orders on the execution path. The present invention retains the expected completion time of the existing work orders on the execution path. The stability of the expected completion time provides a reference for the work efficiency of maintenance personnel, further improving the reliability of the assessment of the surplus threshold.

[0033] (4) The present invention dispatches work orders according to working hours, taking into full consideration the rest periods of maintenance personnel. In addition, in this implementation fee, undispatched work orders are dispatched according to the time of submission, adhering to the principle of first-submitted-first-dispatched, thus avoiding the situation where work orders are squeezed and delayed, and facilitating the timely processing of work orders.

[0034] (5) The present invention provides two methods for inserting work orders into the middle of the execution path. The first method distributes work orders based on the surplus threshold, ensuring the average work saturation of each maintenance personnel; the second method selects the target personnel based on the screening conditions, ensuring the reasonable planning of the execution path of the maintenance personnel, thereby increasing the density of work orders on the execution path, greatly shortening the time the maintenance personnel spend on the road, and improving the efficiency of work order execution.

[0035] (6) In the present invention, when inserting a work order into the middle of the execution path, the reference value is combined to fully consider the impact of the inserted work order on the length of the execution path, so as to insert the work order with the goal of achieving the minimum path length, and to retrieve the travel time of the maintenance personnel as much as possible, thereby improving the ratio of the actual processing time of the work order to the travel time, which is conducive to significantly improving the work order processing efficiency.

[0036] (7) In the present invention, the conditions F1 and F2 are combined to determine whether the work order is inserted into the middle position or the end position of the execution path, thereby avoiding the delay of the existing work orders on the execution path caused by the insertion of the work order. While ensuring the efficiency of work order dispatching, it also ensures the efficiency of work order processing.

[0037] (8) The present invention realizes the timely recovery of unprocessed work orders, avoids the work orders that have been dispatched from being put on hold due to reasons of maintenance personnel, and avoids the impact of personal reasons on the speed of work order processing. While ensuring the overall work order processing efficiency, it also avoids the situation where individual work orders are delayed for a long time.

[0038] (9) The power marketing work order dispatching system and storage provided by the present invention provide a carrier for the above-mentioned power marketing work order dispatching method, which is conducive to the promotion and application of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The following is a flow chart of a method for dispatching work orders for power marketing. DETAILED DESCRIPTION

[0040] Route: The traffic route passing through a set location can be planned based on an electronic map using an existing navigation system.

[0041] Reference Figure 1 , a method for dispatching power marketing work orders proposed in this embodiment includes the following steps S1 to S4.

[0042] S1. Set the work order category and job type, with a one-to-one correspondence between the work order category and the job type; the job type includes the professional direction and level of the maintenance personnel; set up a work order classification engine, which is used to classify the obtained work orders; each work order category is associated with a corresponding expected time.

[0043] In power marketing operations, different maintenance personnel have different operating permissions for the power network. Therefore, when a work order is obtained, the operating permissions required for the work order must be determined first before the work order can be assigned. If the permissions of the person receiving the work order are insufficient, the corresponding tool application permissions and power network component operating permissions cannot be obtained, and thus the work order cannot be executed.

[0044] In this step, by mapping work order categories to job types, maintenance personnel are divided into different job types based on their expertise and level. The expertise indicates the maintenance personnel's technical expertise, while the level specifies their permissions, including the range of tools they can access and their authority to operate power network components. This categorization of work orders by job type ensures that each work order is assigned to the appropriate maintenance personnel for smooth execution.

[0045] In this step, the work order classification engine is implemented using existing technical means, such as using keyword extraction technology, semantic recognition technology, etc. to classify work order keywords. The existing neural network model can also be used to generate a work order classification engine by autonomously learning manually labeled work orders.

[0046] S2. Set the working period and execution path. The working period is a continuous time period. The execution path corresponds to the maintenance personnel one-to-one. The execution path is a traffic path starting from the current location of the maintenance personnel and passing through the locations where the maintenance personnel report unfinished work orders. Each work order on the execution path is associated with an expected completion time; set the surplus weight a, a=b1×b2+c1×c2+d1×d2; b1 is the number of unfinished work orders, b2 is the weight of b1; c1 is the actual completion time Tl' of the most recently completed work order minus the expected completion time Tl of the work order, that is, c1=Tl'-Tl; c2 is the weight of c1; d1 is the absolute value of the difference between the current time T0 and the end time Tn of the current execution path, that is, d1=|T0-Tn|; d2 is the weight of d1; b2, c2 and d2 are all set values; the end time of the current execution path is the expected completion time of the last work order on the current execution path;

[0047] In this step, b1 reflects the maintenance personnel's task progress, c1 reflects the maintenance personnel's work order completion speed, and d1 reflects the remaining working hours in the current working period. The surplus weight a calculated by combining the maintenance personnel's task progress, speed, and remaining working hours can accurately assess the saturation level of the maintenance personnel's current work order allocation, making it easier to subsequently dispatch work orders based on the maintenance personnel's work saturation, thereby ensuring that each work order can be processed in a timely manner.

[0048] In this embodiment, work orders are allocated for each work order category. The work order distribution for a single work order category specifically includes the following steps S3-S4:

[0049] S3. At the start of each work period, select the n(k) work orders with the earliest reporting time from the unassigned work orders of category I(k) as the starting work orders. Assign these n(k) starting work orders to each maintenance worker in the job category corresponding to the work order of category I(k), where n(k) is the total number of maintenance workers in the job category corresponding to the work order of category I(k); 1≤k≤K; K is the total number of work order categories, which is also the total number of job categories. Let the maintenance worker's current location be the starting point and the address of the starting work order corresponding to the maintenance worker be the end point. Construct the maintenance worker's execution path. The execution path starts at the maintenance worker's current location and passes through the reported locations of the maintenance worker's assigned but uncompleted work orders. For example, suppose the maintenance worker's current location is E0 and the maintenance worker currently has five uncompleted work orders, reported at the locations of E1, E2, E3, E4, and E5, respectively. The maintenance worker's current execution path is E0-E1-E2-E3-E4-E5.

[0050] S4. For the unassigned work orders of category I(k), select the assigned work orders in descending order of reporting time. For each assigned work order, first determine whether there is a maintenance personnel who meets both the following conditions F1 and F2:

[0051] F1: The length of the shortest path between the reported location of the dispatched object and the execution path of the maintenance personnel is less than or equal to the set path threshold;

[0052] F2: Among the maintenance personnel who meet the condition F1, there are maintenance personnel whose surplus weight is greater than or equal to the set surplus threshold;

[0053] If there are maintenance personnel who meet the above conditions F1 and F2, the dispatch object is inserted into the execution path of one of the maintenance personnel who meet the conditions, specifically, the following steps S40 can be performed;

[0054] If there is no maintenance personnel who meets the above conditions F1 and F2, the following steps S41-S42 are executed.

[0055] Each work order on the execution path is associated with an expected completion time. The expected completion time of the starting work order is the start time of the corresponding work period plus the expected time corresponding to the starting work order; the expected completion time of the dispatch object inserted between any two work orders on the execution path is the expected completion time of the next work order of the dispatch object on the execution path minus half of the expected time corresponding to the dispatch object; the expected completion time of the dispatch object as the end point of the execution path is the expected completion time of the previous work order of the dispatch object plus the expected time corresponding to the dispatch object.

[0056] S40: Select the maintenance personnel with the largest corresponding surplus weight a from the maintenance personnel who meet the conditions F1 and F2 as the target personnel, and insert the dispatch object into the execution path of the target personnel; or record the maintenance personnel who meet the conditions F1 and F2 as alternative personnel, and record the execution path of the alternative personnel as the alternative path. For each alternative path, calculate the path length between the reporting address of the dispatch object and the reporting addresses of each work order on the alternative path as the reference value, calculate the sum of the reference values ​​corresponding to the two adjacent work order reporting addresses on the alternative path, and select the minimum sum of the reference values ​​as the alternative path for the filtering condition value of the dispatch object. Select the alternative path with the minimum filtering condition value as the target path, assign the dispatch object to the maintenance personnel corresponding to the target path, i.e. the target personnel, and insert the reporting address of the dispatch object between the two work order reporting addresses corresponding to the minimum sum of the reference values ​​on the target path to form a new execution path.

[0057] In this step, when selecting target personnel based on the surplus weight a, the average work saturation of each maintenance personnel is guaranteed; when selecting target personnel based on the screening conditions, the reasonable planning of the maintenance personnel's execution path is guaranteed, thereby increasing the density of work orders on the execution path, greatly shortening the time maintenance personnel spend on the road, and improving the efficiency of work order execution.

[0058] Specifically, when the maintenance personnel with the largest surplus weight a is selected as the target personnel based on the surplus weight a, the execution path of the target personnel is used as the target path, and the path length between the reporting address of the dispatch object and the reporting addresses of each work order on the target path can be further calculated as a reference value, and the sum of the reference values ​​corresponding to the two adjacent work order reporting addresses on the target path is calculated, and the reporting address of the dispatch object is inserted between the two work order reporting addresses corresponding to the minimum sum of the reference values ​​on the target path to form a new execution path.

[0059] S41. Obtain the end time corresponding to each current execution path. The end time is calculated as follows:

[0060] Ts(k)=T0+n'(k)×T(k);

[0061] Where Ts(k) is the end time of any execution path in the work type corresponding to the I(k)th work order, T0 is the current time, n'(k) is the number of work orders remaining on the execution path, and T(k) is the expected time associated with the I(k)th work order;

[0062] S42. Determine whether there is an execution path that satisfies Ts(k)+T(k) less than or equal to the end point of the current work period; if not, wait for the start point of the next work period and assign the dispatch object as an undispatched work order;

[0063] If yes, then obtain all execution paths that satisfy Ts(k)+T(k) less than or equal to the end point of the current working period as alternative paths, calculate the path length between the end point of each alternative path and the reported address of the dispatch object, select the alternative path with the smallest corresponding path length as the target path, assign the dispatch object to the maintenance personnel corresponding to the target path, and update the end point of the target path to the reported address of the dispatch object.

[0064] In this implementation, work orders are assigned based on working hours, fully taking into account the maintenance personnel's rest periods. Furthermore, unassigned work orders are assigned based on the time they were submitted, adhering to the principle of first-submitted, first-assigned. This avoids delays in work orders and facilitates their timely processing.

[0065] S5. Collect the unfinished work orders on each execution path, and at the starting point of the next work period, distribute the collected unfinished work orders as undispatched work orders.

[0066] Through step S5, the unprocessed work orders are recovered in time, which avoids the work orders that have been dispatched from being put on hold due to maintenance personnel's reasons, and avoids the impact of personal reasons on the work order processing speed. While ensuring the overall work order processing efficiency, it also avoids the situation where individual work orders are delayed for a long time.

[0067] The following describes the above work order dispatching method using a specific scenario. Example

[0068] The expected duration of a work order of type X is Tx. During a certain work period, there are two maintenance personnel on duty for the work type corresponding to the work order of type X: Maintenance Personnel A and Maintenance Personnel B. Let Maintenance Personnel A's execution path be Execution Path A, and Maintenance Personnel B's execution path be Execution Path B.

[0069] The unfinished work orders on execution path A are: Fa1, Fa2, Fa3, Fa4;

[0070] The unfinished work orders on execution path B are: Fb1, Fb2, Fb3;

[0071] The end time Ts of execution path A 甲 =T0+4×Tx;

[0072] End time Ts of execution path B 乙 =T0+3×Tx;

[0073] Ts 甲 and Ts 乙 The path length from the end point of the current working period is greater than Tx.

[0074] The current surplus weight of maintenance worker A is a 甲 Greater than the surplus weight a of maintenance worker B 乙 .

[0075] Let the path threshold be L0 and the surplus threshold be a0.

[0076] Among the currently undispatched work orders, the earliest reported time is Fc0.

[0077] Example 1

[0078] In the context of the above embodiment, it is assumed that a 甲 and a 乙 Both are greater than a0, and the length of the shortest path between the reported address of Fc0 and execution path A is less than L0, and the length of the shortest path between the reported address of Fc0 and execution path B is less than L0. That is, for the unassigned work order Fc0, maintenance personnel A and B both meet conditions F1 and F2.

[0079] At this point, there are two ways to arrange the undispatched work order Fc0.

[0080] The first method: select the target path based on the surplus threshold and insert the undispatched work order Fc0

[0081] Assume a 甲 Greater than a 乙 , then select execution path A as the target path and further calculate the following path lengths:

[0082] La(1,c,2)——represents the path length from the address reported by Fa1 to the address reported by Fa2 via the address reported by Fco;

[0083] La(2,c,3)——represents the path length from the address reported by Fa2 to the address reported by Fco to the address reported by Fa3;

[0084] La(3,c,4)——represents the path length from the address reported by Fa3 to the address reported by Fco to the address reported by Fa4;

[0085] Let La(2,c,3) be less than La(1,c,2) and La(3,c,4), then insert Fco between Fa2 and Fa3, that is, work order Fco is assigned to maintenance personnel A, and the unfinished work orders on execution path A are updated in sequence: Fa1, Fa2, Fco, Fa3, Fa4.

[0086] The second method: select the target path based on the distance and insert the undispatched work order Fc0

[0087] In this embodiment, La(1,c,2), La(2,c,3), La(3,c,4), Lb(1,c,2), and Lb(2,c,3) need to be calculated. The definitions of La(1,c,2), La(2,c,3), and La(3,c,4) refer to the first method.

[0088] Lb(1,c,2)——represents the path length from the address reported by Fb1 to the address reported by Fb2 via the address reported by Fco;

[0089] Lb(2,c,3)——represents the path length from the address reported by Fb2 to the address reported by Fb3 via the address reported by Fco;

[0090] Let Lb(1,c,2) be less than La(1,c,2), La(2,c,3), La(3,c,4) and Lb(2,c,3), then select execution path B as the target path, insert Fco between Fb1 and Fb2, that is, work order Fco is assigned to maintenance personnel B, and the unfinished work orders on execution path B are updated in sequence: Fb1, Fco, Fb2, Fb3.

[0091] Example 2

[0092] In the context of the above embodiment, assuming that maintenance personnel A and B cannot simultaneously meet the above conditions F1 and F2, and the path length between the Fa4 reporting address and the Fco reporting address is Lac, and the path length between the Fb3 reporting address and the Fco reporting address is Lbc, and Lac is less than Lbc, then Fco is assigned to maintenance personnel A, and Fco is inserted after Fa4, that is, the unfinished work orders on execution path A are updated to: Fa1, Fa2, Fa3, Fa4, Fco.

[0093] In this embodiment, two specific implementation methods of the work order classification engine are also provided.

[0094] The first work order classification engine

[0095] Keyword extraction technology, semantic recognition technology, etc. are used to classify work orders according to their keywords. That is, each work order category is associated with a corresponding keyword set. The work order classification engine is used to extract the keywords of each work order and calculate the overlap between the keywords of the work order and the keywords in each keyword set. The work order category corresponding to the maximum overlap is selected as the category to which the work order belongs.

[0096] The second work order classification engine

[0097] An existing neural network model is used to autonomously learn from manually annotated work orders to generate a work order classification engine. Specifically, work types and their corresponding work order categories are first set. Historical work orders are then manually annotated to form training samples. The initialized neural network model learns from these training samples to update its parameters. The learned neural network model is then used as the work order classification engine.

[0098] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for dispatching power marketing work orders, characterized in that: The following steps are involved: S1. Set work order categories and job types, with a one-to-one correspondence between work order categories and job types. Job types include the professional field and level of maintenance personnel. Set up a work order classification engine, which is used to classify received work orders. Each work order category is associated with a corresponding expected time. S2. Set the working period and execution path. The working period is a continuous time period. The execution path corresponds to the maintenance personnel one-to-one. The execution path is a traffic path starting from the maintenance personnel's current location and passing through the locations where the maintenance personnel reported unfinished work orders. Each work order on the execution path is associated with an expected completion time. Set the surplus weight a, a = b1×b2+c1×c2+d1×d2; b1 is the number of unfinished maintenance work orders, c1 is the actual completion time of the most recently completed work order minus the expected completion time, and d1 is the absolute value of the difference between the current time and the end time of the current execution path; b2, c2, and d2 are weights, and are all set values; the end time of the current execution path is the expected completion time of the last work order on the current execution path; Work orders are assigned for each work order category. The work order distribution for a single work order category specifically includes the following steps S3-S4: S3. At the starting point of each work period, select the n(k) work orders with the earliest reporting time from the unassigned work orders of the I(k)th work order as the starting work order, and assign the n(k) starting work orders to each maintenance personnel of the work type corresponding to the I(k)th work order, where n(k) is the total number of on-duty maintenance personnel of the work type corresponding to the I(k)th work order; 1≤k≤K; K is the total number of work order categories set manually, which is also the total number of work types; let the maintenance personnel's current location be the starting point and the report address of the maintenance personnel's corresponding starting work order be the end point to construct the maintenance personnel's execution path; S4. For the undispatched work orders of category I(k), select the undispatched work orders as dispatch targets in descending order of reporting time. For each dispatch target, first determine whether there is a maintenance personnel who satisfies both the following conditions F1 and F2: F1: The length of the shortest path between the reported location of the dispatched object and the execution path of the maintenance personnel is less than or equal to the set path threshold; F2: Among the maintenance personnel who meet the condition F1, there are maintenance personnel whose surplus weight is greater than or equal to the set surplus threshold; If there is a maintenance personnel who meets the above conditions F1 and F2, the dispatch object is inserted into the execution path of one of the maintenance personnel who meets the conditions; If there is no maintenance personnel who meets the above conditions F1 and F2, the dispatch object will be assigned to a maintenance personnel of the corresponding work type and used as the end point of the maintenance personnel's execution path, or wait for the starting point of the next work period and assign the dispatch object as an undispatched work order; Each work order on the execution path is associated with an expected completion time. The expected completion time of the starting work order is the start time of the corresponding work period plus the expected time corresponding to the starting work order; the expected completion time of the dispatch object inserted between any two work orders on the execution path is the expected completion time of the next work order of the dispatch object on the execution path minus half of the expected time corresponding to the dispatch object; the expected completion time of the dispatch object as the end point of the execution path is the expected completion time of the previous work order of the dispatch object plus the expected time corresponding to the dispatch object.

2. The power marketing work order dispatching method according to claim 1, characterized in that: In S4, if there are maintenance personnel who meet the above conditions F1 and F2, the maintenance personnel with the largest corresponding surplus weight a is selected as the target personnel from the maintenance personnel who meet the conditions F1 and F2, and the dispatch object is inserted into the execution path of the target personnel.

3. The power marketing work order dispatching method according to claim 2, characterized in that: The method for inserting the dispatch object into the execution path of the target personnel is as follows: calculate the path length between the reporting address of the dispatch object and the reporting addresses of each work order on the current execution path of the target personnel as a reference value, calculate the sum of the reference values ​​corresponding to the two adjacent work order reporting addresses on the current execution path of the target personnel, obtain the two work orders corresponding to the minimum sum of the reference values ​​as reference objects, and insert the dispatch object between the two reference objects to obtain a new execution path.

4. The power marketing work order dispatching method according to claim 1, characterized in that: In S4, if there are maintenance personnel who meet the above conditions F1 and F2, the maintenance personnel who meet the conditions F1 and F2 will be recorded as alternative personnel, and the execution path of the alternative personnel will be recorded as the alternative path. For each alternative path, the path length between the reporting address of the dispatch object and the reporting addresses of each work order on the alternative path is calculated as the reference value, the sum of the reference values ​​corresponding to the two adjacent work order reporting addresses on the alternative path is calculated, and the minimum sum of the reference values ​​is selected as the filtering condition value for the dispatch object. The alternative path with the minimum filtering condition value is selected as the target path, the dispatch object is assigned to the maintenance personnel corresponding to the target path, and the reporting address of the dispatch object is inserted between the two work order reporting addresses corresponding to the minimum sum of the reference values ​​on the target path to form a new execution path.

5. The power marketing work order dispatching method according to claim 1, characterized in that: In S4, if there is no maintenance personnel meeting the above conditions F1 and F2 for the dispatch object, the following sub-steps are executed: S41. Obtain the end time corresponding to each current execution path. The end time is calculated as follows: Ts(k)=T0+n'(k)×T(k); Where Ts(k) is the end time of any execution path in the work type corresponding to the I(k)th work order, T0 is the current time, n'(k) is the number of work orders remaining on the execution path, and T(k) is the expected time associated with the I(k)th work order; S42. Determine whether there is an execution path where Ts(k)+T(k) is less than or equal to the end point of the current work period; if not, wait until the start point of the next work period and assign the dispatch object as an undispatched work order; If yes, then obtain all execution paths that satisfy Ts(k)+T(k) less than or equal to the end point of the current working period as alternative paths, calculate the path length between the end point of each alternative path and the reported address of the dispatch object, select the alternative path with the smallest corresponding path length as the target path, assign the dispatch object to the maintenance personnel corresponding to the target path, and update the end point of the target path to the reported address of the dispatch object.

6. The power marketing work order dispatching method according to claim 1, characterized in that: At the end of each work period, step S5 is executed: unfinished work orders on each execution path are collected, and at the starting point of the next work period, the collected unfinished work orders are distributed as undispatched work orders.

7. A power marketing work order dispatching system, characterized in that: It includes a memory and a processor, the memory stores a computer program, the processor is connected to the memory, and the processor is used to execute the computer program to implement the power marketing work order dispatching method as described in any one of claims 1 to 6.

8. A storage medium, characterized in that: A computer program is stored, and when the computer program is executed, it is used to implement the power marketing work order dispatching method according to any one of claims 1 to 6.

Images