JP4894035B2 - On-site work planning system, method and program - Google Patents

Description

  The present invention relates to a maintenance work plan for trains (train formation) at railway vehicle bases, and an on-site work plan system, method, and program for supporting creation of a train replacement plan within a station premises by a computer.

  In the premises work at the depot, the incoming vehicles are drawn into the inspection number line, cleaning number line, indwelling number line, etc. in knitting units, the work is carried out according to the respective restrictive conditions, and then departed. The formation refers to a plurality of vehicles that move together (connected to each other). This on-site work plan is a type of problem that is very difficult to apply general optimization methods, and it is difficult to respond to unique special constraints that differ depending on the vehicle base, type of vehicle organization, and driver's work form, At the present time, veteran staffs are planning through trial and error.

  As a conventional method, a method like the following patent document has been proposed.

  The method disclosed in Japanese Patent Laid-Open No. 2000-190849 limits the typical replacement / work process from entry to departure to “departure → detainment → work → detention → department” and has a predetermined organization. According to the order, the use number line of the work is determined and then the detention number line and the time are determined repeatedly.

  The method of Japanese Patent Laid-Open No. 2000-177590 limits the problem to the exchange plan creation in the station premises, and expresses the plan as network format data called PERT. Specifically, for each train, a network is created in which nodes representing arrival at each track or departure from each track are connected according to the execution order by arcs weighted by the required time. Then, a plan that satisfies all the constraints is created by repeatedly changing and evaluating the arc connection relationship corresponding to the competition order of the work processes competing for resources (numbered lines) and performing a search.

The method disclosed in Japanese Patent Application Laid-Open No. 2005-178742 formulates a work process in a form called “replacement sequence” and then formulates it as a mathematical model in a form that can be processed by constraint programming to create a provisional solution of the plan. If no solution satisfying the constraint is found, a part of the replacement sequence is automatically changed and the provisional solution is found again.
JP 2000-177590 A JP 2005-178742 A JP 2000-190849

  Conventionally proposed on-site work plan automatic proposal methods are difficult to achieve both the versatility of the method and the scalability of performance.

  For example, the method disclosed in Japanese Patent Laid-Open No. 2000-190849 does not use only the typical replacement / work process described above in an actual vehicle depot, but waits while detouring across a plurality of detention lines. It is necessary to use various procedures depending on the situation, such as carrying out the work as it is with the retained wire, but this method is expected to be difficult to handle and lacks versatility.

  In addition, the method disclosed in Japanese Patent Application Laid-Open No. 2000-177590 uses the method of changing the arc connection relationship in the station premises (upline change, two-stage replacement, change of the order of use of the pulling line, change of course point use order) ), And is a search method targeting a specific work process in the station premises, which also lacks versatility.

  On the other hand, in the method of Japanese Patent Application Laid-Open No. 2005-178742, in the basic solution of constraint programming, when the number of organizations is large and mutual interference increases (congests), considerable computation is required to solve each mathematical model. Although it is expected to take time, in an actual depot, the success or failure of the work plan is often determined by a combination of work processes (replacement sequence). If the base is crowded with a large number of trains, it is expected that it will take a considerable amount of computing time to optimize including the combination of work processes (replacement sequences), which will improve performance scalability. Lack.

  The present invention provides an on-site work planning system, method, and program for automatically proposing a on-site work plan with emphasis on practicality that can achieve both versatility and performance scalability.

  An on-site work plan creation system as one aspect of the present invention creates a on-site work plan for creating a on-site work plan for the formation of a vehicle from when the vehicle formation enters the depot to when the work is completed and exited. A system for storing a plurality of basic information storage means for storing basic information necessary for a work plan of each organization and a business template representing a network in which nodes indicating the start or end of a work process are connected according to an execution order by an arc The arc is weighted by a minimum required time, a maximum required time, or both, and a set of resource candidates that can be used in each work process in the business template is defined for each business template. Resource definition data storage means for storing resource definition data to be processed, and the business template Business template selection means for selecting a business template for the target organization from the storage means, upper limit time and lower limit time of each node in the network represented by the selected business template, basic information on the target organization, and A network time calculation means for calculating from the arc weight value between the nodes, and whether or not the network represented by the selected business template can be integrated into a local work network representing a local work plan. Inspecting based on the resource definition data corresponding to the network and the selected business template, and if possible, network sequential update means for integrating the network represented by the selected business template into the local work network, The network time calculation means Used, comprising the recalculates the upper time and lower time of each node in the yard work network, and updates the upper limit time and lower time of each node by the value calculated.

  According to another aspect of the present invention, there is provided an on-site work plan creation method for creating an on-site work plan for forming a vehicle from when the vehicle formation enters the vehicle base to when the vehicle finishes and exits. A basic information database for storing basic information necessary for a work plan for each organization, and a plurality of business templates representing a network in which nodes indicating the start or end of a work process are connected according to an execution order by an arc. The arc is weighted by a minimum required time, a maximum required time, or both, and a resource that defines a set of resource candidates that can be used in each work process in the business template for each business template. Preparing a resource definition database for storing definition data; A business template selection step for selecting a business template for the target organization from the business template database, and the upper limit time and the lower limit time of each node in the network represented by the selected business template, and the basic information of the target organization A network time calculation step for calculating from the arc weight value between the nodes, and whether or not the network represented by the selected business template can be integrated into a local work network representing a local work plan. A network sequential update step for examining based on the resource definition data corresponding to the local work network and the selected business template and, if possible, integrating the network represented by the selected business template into the local work network; The recalculate the upper time and lower time of each node in the yard work network, and a network time update step of updating the upper limit time and lower time of each node by calculation value, characterized by including the.

  The on-site work plan creation program as one aspect of the present invention is a premise work for creating a on-site work plan for the formation of the vehicle from the time the vehicle formation enters the vehicle base to the time when the work is completed and exited. A plan creation program representing a network in which a basic information database storing basic information necessary for a work plan of each organization is connected, and a node indicating the start or end of a work process is connected according to an execution order by an arc A plurality of business templates are stored, the arc is weighted by a minimum required time, a maximum required time or both, and a step of accessing a business template database, and each business template is used in each work process in the business template Resource definition data that defines a set of possible resource candidates A resource definition database that stores information, a business template selection step that selects a business template for a target organization from the business template database, and an upper limit time and a lower limit of each node in the network represented by the selected business template A network time calculating step for calculating time from basic information of the target organization and a weight value of an arc between the nodes, and a network operation represented by the selected business template, Whether or not the network can be integrated is checked based on the resource definition data corresponding to the local work network and the selected business template, and if possible, the network represented by the selected business template Network sequential update step for integrating into the local work network, and network time update step for recalculating the upper limit time and lower limit time of each node in the local work network and updating the upper limit time and lower limit time of each node with the calculated values And make the computer execute.

  According to the present invention, it is possible to automatically propose a premises work plan that emphasizes practicality and can achieve both versatility and performance scalability.

  In the present embodiment, one or a plurality of business templates representing work processes that can be selected for a train (vehicle organization) and candidates for time constraints that must be observed during the work process are prepared for each train, and a work plan on the premises is prepared. When creating a business plan, the business templates determined for each train are integrated into one plan. The parts depending on each vehicle base can be dealt with only by changing the business template to realize versatility. Hereinafter, this embodiment will be described with reference to the drawings.

  FIG. 1 is a diagram showing a configuration of an embodiment of a site work planning system according to the present invention.

  The entire system includes a business template input / display unit 1, a schedule display unit 2, a data storage unit 3, and a network sequential update unit 4. The function by each means may be realized by causing a computer to execute a program, or may be realized by hardware.

  The business template input / display unit 1 is an interface for a user to register a business template and input various instructions.

  The data storage means 3 is a database for storing necessary data. The business template storage means (business template database) 30, the resource definition data storage means (resource definition database) 31, the composition information / basic diamond storage means (basic information) Storage means or basic information database) 32 and schedule storage means 33. The business template storage unit 30 stores a plurality of business templates. The resource definition data storage unit 31 stores resource definition data. The organization information / basic schedule storage means 32 has basic information (organization ID, organization type, scheduled entry / exit times) necessary for the on-site work plan for each organization that has been determined to enter and exit the depot and station. , To store available business templates and resource definition data specification information). The schedule storage means 33 stores the on-site work plan (schedule or on-site work network) created by the network sequential update means 4.

  The network sequential update unit 4 includes a work process pair node insertion unit 40, a network time calculation unit 41, and a network temporary storage unit 42. The network sequential update means 4 determines a work plan for each composition in the composition information / basic diagram storage means 32 and creates a local work plan (schedule or local work network) by integrating the work plans for each composition. The business template selection means 5 determines the train to be added next from the scheduling information / basic schedule storage means 32 based on the set priority, and determines the available business templates for the determined train from the business template storage means 30. The resource definition data associated with the business template is extracted from the resource definition data storage unit 31 and passed to the network sequential update unit 4. The business template data and resource definition data will be described later.

  The schedule display means 2 is an interface for confirming the on-site work plan (schedule or on-site work network) created by the network sequential update means 4.

  In the following, the outline of PERT, business template, and resource definition data will be described first.

  The automatic proposal of on-site work plans has many similarities to the project scheduling problem. In the present embodiment, the business template is described in a format conforming to PERT (Project Evaluation and Review Technique), which is a technique having a track record in the field of project management. In PERT, a plan is expressed as a network structure called an arrow graph (or arrow diagram). One work process is represented as an arc, and both ends of the arc are represented as nodes indicating the start and end of the work process, and connected according to the temporal causal relationship of the work process. If all work processes are organized and described according to such a notation, it becomes possible to estimate the deadline, start date, margin time, etc. for each job by a simple iterative calculation process.

Only the relevant part of PERT will be explained below using an example. Assume that the number of days required for the work of W1 to W8 and the context between the works are given as shown in the table below. When there is a work that cannot be started unless a certain work is completed, this certain work is called a preceding work. For example, W4 cannot start unless W1 is completed, and W1 is a preceding work of W4. W6 cannot be started unless all of W2, W3, and W4 are completed, and W2, W3, and W4 are the preceding operations of W6.

  In PERT, first, a network as shown in FIG. 2 is created.

  In PERT, a work is represented by an arrow, which is represented as an activity, and its starting point / completion point is represented as a node, which is an event. Number the events so that they do not overlap. Multiple activities may flow into one event or multiple activities may flow out from one event. For example, event 4 in FIG. 1 is also the point where W3 and W4 are completed, and the point where W6 and W7 start But there is.

Then, to complete this project in the shortest time, ask what work to start from and when to complete it. Here, the time required for the work ij is represented by Tij. The time at which all of the activities (work) flowing out from the event i can be started earliest is referred to as the earliest connection point time (ETi). First, the project start time (ET1) is set to day 0. Then, since the required time T12 of the work 1-2 is 6, the work starting from the event 2 is 6 days at the earliest. That is, the earliest connection point time (ET2) of event 2 is 6 days. Similarly, in event 3, ET3 = ET1 + T13 = 0 + 7 = 7 days. There are 3 activities flowing in at Event 4,

Work 1-4 completion date: ET1 + T14 = 0 + 5 = 5 days Work 2-4 completion date: ET2 + T24 = 6 + 4 = 10 days Work 3-4 completion date: ET3 + T34 = 7 + 0 = 7 days

However, the work after event 4 can be performed after all of these work is completed, so ET4 becomes the maximum value of these 10 days. These can be summarized as the following formula.

ET1 = 0
ETj = max {ETh + Thj}
(H is the start event of work that flows into event i)
For example, in event 5,
ET5 = max {(ET3 + T35), (ET4 + T45)}
= Max {(7 + 3), (10 + 8)}
= Max {10, 18}
= 18

It becomes. Thus, when the ETi (earliest connection point time) is calculated for all events, it is as shown in FIG. 3, and it can be seen from ET6 that the shortest completion time of this project is 27 days.

  Next, the latest connection point time is the time that all work that flows into event j must be completed by this time at the latest in order to complete the project with the shortest completion time. Indicated by LTj. This is calculated sequentially from the last event of the project. If the project deadline is within 37 days, LT6 will be 37 days.

The calculation of LTj is expressed by the following equation.

LTn = ETn (n: final event)
LTi = min {LTk−Tik} (k: event of work flowing out from i)

For example, in event 4, the following occurs.

LT4 = min {(28-8), (37-10)} = 20

FIG. 3 shows the LTi (latest connection time) for all events.

  Next, the business template will be described.

  In the business template, the composition movement / detention, inspection, and cleaning work are regarded as arcs, and the start time and the end time of the work are respectively connected as the start / end event nodes in the order of execution. Information indicating a minimum time interval and a maximum time interval is added to the arc as a time constraint to be observed between events. Further, information indicating the earliest execution time (lower limit time) restriction and the latest execution time (upper limit time) restriction that can be executed as a restriction of an event alone is added to the node. The minimum time interval, the maximum time interval, the earliest execution time limit, and the latest execution time limit may be omitted.

  The business templates that can be used for each organization are determined in advance according to the combination of the types of vehicles constituting the organization and the assigned maintenance work. The user registers these business templates in advance as basic data in association with each organization.

  FIG. 4 shows an example of a business template. The business template is internally represented as network structure data.

  In this example, the train enters from the main line, passes through the point and arrives at the arrival and departure line, waits for a while, passes through the point, enters the garage number line, performs maintenance work while waiting for the garage, and then passes the point Re-enter the arrival / departure line, wait for a while, pass through the point and appear on the main line.

  The arc between nodes 0 and 1 has a minimum and maximum time interval of 1, which indicates that the point travel time from the main line to the departure and arrival line is fixed at 1.

  The arc between nodes 0 and 2 has a minimum and maximum time interval of 0, respectively, which indicates a constraint that point movement and incoming line placement must be started simultaneously. In normal train control, safety is emphasized, and control called “blocking” is performed to lock not only the number of the current train but also the number of front and rear lines related to movement, but the above restrictions correspond to this control. It is.

  The arc between nodes 2 and 3 has a minimum time interval of 10, which indicates that a detention time of at least 10 hours must be secured.

  The minimum and maximum time intervals of the arcs between the nodes 4 and 5 are 1, respectively, which indicates that the point movement time from the departure / arrival line to the garage number is fixed to 1.

  The arc between nodes 3 and 4 has a minimum and maximum time interval of −1, respectively, and the arc between nodes 4 and 6 has a minimum and maximum time interval of 0, respectively. This indicates a restriction that the point movement and the garage number line detention must be started simultaneously one hour before the end of the departure / arrival line detention. This restriction is also a restriction based on the “blocking”.

  A nested structure in which a pair of node 7 and node 8 representing the start and end of maintenance work is sandwiched between a pair of node 6 and node 9 representing the start and end of garage number line detention.

  The arc between nodes 6 and 7 has a minimum time interval of 10, which indicates that a minimum of 10 hours must be secured from the start of use of the garage number line to the start of maintenance work.

  The arc between nodes 7 and 8 has a minimum time interval of 30, indicating that a minimum of 30 times must be secured from the start to the end of the maintenance operation.

  An arc is added between node 2 and node 8 and a maximum time interval constraint 50 is added to indicate a time constraint that must be completed within a certain time from the start (entrance) of the departure / arrival line detention.

  The arc between nodes 8 and 9 has a minimum time interval of 10, which indicates that a minimum of 10 times must be secured from the end of maintenance work to the end of detention of the garage number line.

  The arc between the nodes 10 and 11 has a minimum and maximum time interval of 1, which indicates that the point travel time from the garage number line to the departure / arrival line is fixed at 1.

  The arc between the nodes 9 and 10 has a minimum and maximum time interval of −10, and the arc between the nodes 10 and 12 has a minimum and maximum time interval of 0, respectively. This indicates a restriction that point movement and arrival line detention must be started simultaneously 10 hours before the end of detention of the garage number line. This restriction is also a restriction based on the “blocking”.

  The arc between the nodes 12 and 13 has a minimum time interval of 10, which indicates that a detention time for a departure and arrival line of at least 10 hours must be secured.

  The arc between the nodes 13 and 14 has a minimum and maximum time interval of -1, respectively, which indicates that movement must be started from the arrival and departure line one hour before the end of the departure and arrival line. Further, the minimum and maximum time intervals of the arcs between the nodes 14 and 15 are 1, respectively, which indicates that the point moving time from the departure / arrival line to the main line is fixed to 1. These restrictions are also based on the above “blocking”.

  Next, resource definition data will be described.

  A plurality of options are prepared for resources (number line, driver, work team, etc.) that can be used in each work process of the business template. Further, the resource candidates that can be selected are not necessarily determined individually for each work process, and often have an interdependency determined by the route structure. For example, assuming a route structure as shown in FIG. 5, the points that can be used at the time of movement vary depending on whether the garages 4 and 5 are used or the garage 6 is used. In order to define resource candidates including dependencies in a general format, several resource groups (consisting of one or more resources) that can be used for each work process are specified. Further, the resource definition data is internally represented as network format data indicating the dependency between groups.

  FIG. 6 shows an example of resource definition data. This example is an example when the route structure of FIG. 5 is used and the business template of FIG. 4 is followed. For example, when only the arrival / departure number line 0 can be used, only the points 2, warehouses 4 and 5 can be selected, but when only the arrival / departure number line 1 can be utilized, it can be understood that all garages and points can be selected.

  The business template and resource definition data are associated with each other, and link information is set so that the start node of each work process of the business template and the node indicating the resource group in the resource definition data can be referred to each other Is done. For example, the node 6 indicating the arrival of the garage in the business template of FIG. 4 is associated with two groups of (warehouse 4, warehouse 5) and (warehouse 6) in the resource definition data of FIG.

  Next, the network sequential updating unit 4 will be described.

  The network sequential updating unit 4 integrates business templates registered in individual organization units as one on-site work plan in accordance with the resource competition status. The on-site work plan is also expressed as a network structure. For example, if the route shown in FIG. 7 is followed by the diagram shown in FIG. 8, the on-site work plan (on-site work network) can be represented by an arrow graph as shown in FIG.

  In FIG. 9, a solid line arrow represents a work process for each train (vehicle organization). Work processes using the same resource are arranged in the same horizontal row, and the work processes are connected by dotted arrows. Dotted arrows represent the order in which each train uses its resources. In this example, three dotted arrows Y1 to Y3 are shown. A dotted line arrow Y1 indicates that the train 2 uses the garage 1 after the train 1 leaves the garage 1. A dotted arrow Y2 indicates that the train 2 uses the arrival / departure line 1 after the train 1 leaves the arrival / departure line 1. A dotted arrow Y3 indicates that the train 1 uses the arrival / departure line 1 again after the train 1 leaves the arrival / departure line 1.

  It is not determined in advance which resource each work process uses and in what order. The network sequential updating unit 4 receives network data (see FIG. 4) represented by the business template stored in the business template storage unit 30 for the target train, and simultaneously refers to the resource definition data (see FIG. 6) to be paired. Then, resources are assigned to each work process in the business template. That is, the network represented by the business template is additionally combined with the network representing the on-site work plan (on-site work network). When the work process of the business template causes resource conflict with the work process of the local work network, the work process of the business template is arc-connected to the work process in which resource conflict occurs (see the dotted line arrow in FIG. 9). After network integration, the network time calculation means 41 performs time update calculation of each node in the local work network. Repeat for all trains.

  FIG. 10 is a flowchart schematically showing the flow of processing of the network sequential update means 4.

  First, the train to be added next is determined from the knitting information / basic diamond storage means 32 (S1).

  Next, one business template that can be used for the determined train is extracted from the business template storage means 30, and the resource definition data associated with the business template is extracted from the resource definition data storage means 31 (S2). S2 corresponds to the processing of the business template selection means.

  Based on the extracted business template and resource definition data, an additional calculation is performed on the schedule (premise work plan) of the selected train (S3). Details of S3 will be described later.

  If the addition is successful (YES in S4), it is checked whether the addition has been completed for all trains (S5). If the addition is completed for all trains (YES in S5), the completed premises The work plan (premises work network) is output to the schedule storage means 33. On the other hand, when the train which should be added still remains (NO of S5), it returns to S1 and the train to add next is determined.

  If the addition is not successful (NO in S4), it is checked whether or not the addition calculation has been tried for all business templates that can be used by the selected train (S6).

  If additional operations have not been tried for all available business templates (NO in S6), it is decided to change the business template (S7), and a business template different from the one already tried is taken out (S2). ).

On the other hand, when additional calculations are tried for all available business templates (YES in S6), the process ends if the preset trial end conditions (such as the upper limit of the number of trials and the calculation time limit) are reached (YES in S8). If the trial end condition has not been reached (NO in S8), the priority is changed (for example, the order in which the train is added to the local work network is changed), and the process is restarted from the beginning (S9).
FIG. 11 is a flowchart showing in detail the procedure performed in S3.

When the business template for one organization and the resource definition data paired therewith are extracted from the data storage means 3 in S2 of FIG. 10, a network corresponding to the extracted business template is generated (S11).
A topological sort of the node IDs of the generated network is performed (S12). The topological sort means rearranging the order (ID) so that the arced nodes (upstream side nodes) always have a lower number between the nodes connected by the arc.

  The ET (lower limit time) and LT (upper limit time) of each node in the generated network are calculated using the network time calculation means 41 described later (S13). S13 corresponds to the processing of the network time calculation means.

  A copy of the current on-site work plan (on-site work network) is generated and stored in the network temporary storage means 42 (S14). The current on-site work network is also stored in the network temporary storage means 42.

  Each node in the generated network to be integrated into the current on-site work network is selected in order from the top (S15), and the type of the selected node is checked (S16).

  If the selected node is the end node of the work process (NO in S16), the process proceeds to the next node (S15).

  When the selected node is the start node of the work process (YES in S16), it is checked whether or not resource contention occurs in the work process starting at the start node (S17).

  If no conflict occurs (NO in S17), the node ID of the start node is set as the first node in the resource conflict list management table (S18), and the process moves to the next node (S15).

The resource conflict list management table (see FIG. 12 described later) manages the correspondence between each resource and the resource conflict list. The resource contention list is a part of the premises work network that traces the arc along the nodes that use the specific resource among the nodes of the premises work network, and the work processes that use the specific resource are arranged in the order of use. It corresponds to a thing. In the case of FIG. 9, the resource competition list for the garage 1 is
“First train in 1 garage” → “1 train in 1 garage” → “1 train in 2 garages” → “1 train in 2 garages”
Correspond to a list of nodes (list of nodes in one horizontal row).

  If a conflict occurs in S17 (YES in S17), an optimum insertable location (location where a pair of start node and end node is to be inserted) is searched from the head of the corresponding resource conflict list (S19).

  If the insertion location is not found (NO in S20), the local work network is returned to the one duplicated in S14 (S21), and the process is terminated (composition addition failure). Then, the process proceeds to S4 in FIG.

  If an insertion location is found (YES in S20), a pair of work process nodes is inserted at the corresponding location in the resource conflict list (part of the local work network as described above) (S22).

  Here, the work process pair node insertion means 40 inserts a pair of work process nodes (start node and end node). The insertion method will be described with reference to FIGS. 12 and 13 taking as an example the case where the arrow graph of the train 3 is added to the arrow graph of FIG.

FIG. 12 shows candidates for possible insertion positions of a pair of nodes “3 trains, 1 train” → “3 trains, 1 train”. Although a plurality of cells are shown in the figure, these conceptually show the resource conflict list management table. One square shows the correspondence between one resource and the resource conflict list. In this example, there are three candidate insertion positions (candidates 1 to 3). There are various methods for determining the insertion location, but as a simple method, the insertion is made at the location where the margin time at the time of insertion is considered to be the smallest. The margin time is estimated by the following formula.
Min (LT_next_start − ET_end, LT_start −ET_pre_end)

  Here, LT_start and ET_end are the earliest join point time of the start node and the latest join point time of the end node in the node pair to be inserted, respectively, and ET_pre_end is the earliest join point of the node (end node) located immediately before the insertion point Time, LT_next_start is the latest connection time of the node (start node) located immediately after the insertion point. FIG. 13 shows an example in which the node of “3 trains 1 arrival” → “3 trains 1 departure” is inserted in the candidate 2 location.

  After the addition of the node pair, the topological work network is topologically sorted (S23), and the time update calculation (ET, LT update) is performed using the network time calculation means 41 according to the node ID order (S24). Here, the function of the network time calculation means 41 will be described.

  The time constraint that appears in the campus work plan includes a maximum time interval constraint that is not handled by the conventional PERT. In this embodiment, in order to cope with this maximum time interval constraint, in addition to the time update calculation in the conventional PERT, Add improvements.

  FIG. 14 is a diagram for explaining the time update calculation. Assume that five events (nodes) are shown and all arcs show a minimum time interval 1 and a maximum time interval 2.

  If only the minimum time interval constraint is used, as described in the explanation of PERT, a forward calculation for determining the earliest connection point time ET and a reverse operation for determining the latest connection point time LT are performed once each. Just do it. This is shown in FIG. However, in order to process the maximum time interval constraint, it is necessary to perform the update operation of ET and LT again as shown in FIG. It should be noted that in the processing of FIG. 15B, ET and LT obtained in the first calculation are reused, ET is updated in the backward calculation, and LT is updated in the forward calculation.

  Here, if there is a maximum time interval constraint, mutual contradiction may occur with the minimum time interval constraint. If inappropriate values are set as the maximum time interval and the minimum time interval, the time update operation will not converge. There is a case. For this reason, it is checked whether the set conditions are correct by repeating the prescribed number of forward and reverse operations and confirming whether the values of ET and LT converge.

  If time update calculation (S24) is performed, it will return to S15 and the next node will be selected. If there is no more node to be selected, the process is terminated.

  As described above, according to the present embodiment, it is possible to automatically propose an in-house work plan that emphasizes practicality and can achieve both versatility and performance scalability.

The figure which shows the structure of 1st Embodiment of the premise work planning system by this invention. Diagram showing an example of a network created with PERT The figure which shows the example which added the earliest joining point time and the latest joining point time to the network of FIG. Diagram showing an example of a business template Diagram showing an example of route structure Figure showing an example of resource definition data Diagram showing an example of route structure Diagram showing an example of a train diagram A diagram showing an example of an on-site work plan expressed in an arrow graph The flowchart which shows the flow of a process of a network sequential update means roughly It is a flowchart which shows in detail the procedure performed by S3 of FIG. Figure showing node pair insertion location candidates A diagram showing a node pair inserted Diagram explaining time update calculation Diagram explaining time update calculation

Explanation of symbols

1: Business template input / display unit 2: Schedule display unit 3: Data storage unit 4: Network sequential update unit 30: Business template storage unit 31: Resource definition data storage unit 32: Organization information / basic diagram storage unit 33: Schedule storage Means 40: Work process pair node insertion means 41: Network time calculation means 42: Network temporary storage means

Claims (8)

An on-site work plan creation system for creating an on-site work plan for the formation of the vehicle from the time when the formation of the vehicle enters the depot to the time when the work is finished and the time it leaves.
Basic information storage means for storing basic information necessary for a work plan of each of the above-mentioned organization;
A plurality of business templates representing a network in which nodes indicating the start or end of a work process are connected by arcs in accordance with the execution order are stored. The arcs are weighted with a minimum required time, a maximum required time, or both, and Business template storage means, at least one of which is weighted by the maximum required time ;
Resource definition data storage means for storing resource definition data for defining a set of resource candidates that can be used in each work process in the business template for each business template;
Business template selection means for selecting a business template for the target organization from the business template storage means;
Network time calculation means for calculating the upper limit time and the lower limit time of each node in the network represented by the selected business template from the basic information of the target organization and the weight value of the arc between the nodes;
Inspecting whether the network represented by the selected business template can be integrated into a local work network representing a local work plan based on the resource definition data corresponding to the local work network and the selected business template. And network sequential update means for integrating the network represented by the selected business template into the local work network if possible,
Using the network time calculation means, the upper limit time and the lower limit time of each node in the campus work network are recalculated so as to keep the minimum required time and the maximum required time of the arc, and the upper limit of each node is calculated according to the calculated value. A premises work plan creation system characterized by updating the time and the lower limit time. The network sequential update means includes:
When a resource required by a certain work process in the selected business template conflicts with a resource required by a part of the work process in the campus work plan, another work that competes with the resource in the campus work plan. It is inspected whether or not the certain work process can be added between the processes, and if it can be added, a node pair representing the certain work process is determined as the first work process that is temporally previous among the work processes. Inserted between the node pair representing and the node pair representing the second work process later in time, connecting the start node of the inserted node pair with the end node of the node pair representing the first work process, and inserting 2. The on-site work meter according to claim 1, wherein an end node of the node pair is connected to a start node of the node pair representing the second work process by an arc. Creation system.   The network sequential update means detects between work processes that have the least margin time when a node pair representing a certain work process is inserted, and inserts a node pair representing the certain work process between the detected work processes. The on-site work plan creation system according to claim 2.   The business template selection unit selects another business template that can be used for the target vehicle when the network sequential update unit determines that the selected business template cannot be integrated into the on-site work plan. The on-site work plan creation system according to any one of claims 1 to 3. The network time update means includes
A topological sort based on the arc connection relation of each node in the premises work network is performed, and the upper limit time of each node is calculated in the sort order direction using the minimum required time given to each arc, and is attached to each arc. The lower limit time of each node is calculated in the reverse direction of sorting using the minimum required time, and the lower limit time of each node is calculated in the forward direction of sorting using the maximum required time attached to each arc. 5. The on-site work plan creation system according to claim 1, wherein the upper limit time of each node is calculated in the reverse direction of sorting using the attached maximum required time.   The on-site work plan creation system according to claim 5, wherein the network time update unit repeats the four operations until the upper limit time and the lower limit time at each node converge. An on-site work plan creation method for creating an on-site work plan for the formation of the vehicle from the time when the formation of the vehicle enters the depot to the time when the work is finished and the time it leaves
A basic information database for storing basic information necessary for the work plan of each organization, and a plurality of business templates representing a network in which nodes indicating the start or end of a work process are connected in accordance with the execution order by arcs. A business template database weighted by duration, maximum duration or both and at least one of the arcs is weighted by the maximum duration, and each operation in the business template for each business template Providing a resource definition database storing resource definition data defining a set of resource candidates that can be used in the process;
A business template selection step of selecting a business template for a target vehicle from the business template database;
A network time calculating step of calculating the upper limit time and the lower limit time of each node in the network represented by the selected business template from basic information of the target vehicle and an arc weight value between the nodes;
Inspecting whether the network represented by the selected business template can be integrated into a local work network representing a local work plan based on the resource definition data corresponding to the local work network and the selected business template. And, if possible, a network sequential update step of integrating the network represented by the selected business template into the local work network;
Network time for recalculating the upper limit time and the lower limit time of each node in the premises work network so as to keep the minimum required time and the maximum required time of the arc, and updating the upper limit time and the lower limit time of each node with the calculated values An update step;
A yard work plan creation method characterized by comprising: An on-site work plan creation program for creating an on-site work plan for the formation of the vehicle from the time when the formation of the vehicle enters the depot to the time when the work is finished and exited,
Accessing a basic information database storing basic information necessary for a work plan for each said organization;
A plurality of business templates representing a network in which nodes indicating the start or end of a work process are connected by arcs in accordance with the execution order are stored. The arcs are weighted with a minimum required time, a maximum required time, or both, and Accessing a business template database, at least one of which is weighted by the maximum required time ;
Accessing a resource definition database that stores resource definition data that defines a set of resource candidates that can be used in each work process in the business template for each business template;
A business template selection step of selecting a business template for a target vehicle from the business template database;
A network time calculating step of calculating the upper limit time and the lower limit time of each node in the network represented by the selected business template from basic information of the target vehicle and an arc weight value between the nodes;
Inspecting whether the network represented by the selected business template can be integrated into a local work network representing a local work plan based on the resource definition data corresponding to the local work network and the selected business template. And, if possible, a network sequential update step of integrating the network represented by the selected business template into the local work network;
Network time for recalculating the upper limit time and the lower limit time of each node in the premises work network so as to keep the minimum required time and the maximum required time of the arc, and updating the upper limit time and the lower limit time of each node with the calculated values An update step;
On-site work plan creation program to make computer execute.

Images