A method and system for routing optimization planning of a triple play engineering optical cable

Abstract

The application discloses a triple-play engineering optical cable routing optimization planning method and system, and the system comprises the following steps: constructing a network topology graph, including constructing a network topology graph according to source nodes, target nodes and path vector data; completing constraint condition quantization, including extracting constraint parameters from engineering specification data, calculating constraint weights of each edge in the network topology graph according to the constraint parameters, and obtaining comprehensive cost weights; performing path optimization, including searching candidate paths from the source nodes to each target node based on the comprehensive cost weights by using a graph search algorithm; performing multi-dimensional evaluation on the candidate paths to obtain final planning paths; and optimizing resource allocation, including calculating fiber core number demand quantity layer by layer according to the final planning paths and target node coverage household numbers, and selecting capacity specifications meeting the demand from wiring equipment specification sequences. The application can effectively improve the scientificity of triple-play engineering optical cable routing planning and the rationality of fiber resource allocation.

Description

Technical Field

[0001] This application relates to the field of engineering planning technology, specifically to a method and system for optimizing optical cable routing in a three-network convergence project. Background Technology

[0002] With the rapid development of communication technology, the convergence of telecommunications networks, broadcast television networks, and the Internet has become an important trend in modern building communication infrastructure. The "three-network convergence" project enables unified planning and laying of optical cables for various communication services, achieving intensive utilization of communication resources. In the optical cable routing planning of building complexes or residential communities, optical cables need to be laid from the communication equipment room within the site to the fiber optic distribution boxes on each floor, involving various laying environments such as underground pipelines, vertical shafts, and horizontal passageways. The quality of optical cable routing planning has a direct impact on project cost, construction difficulty, and subsequent maintenance efficiency. In today's digital and information-driven era, efficient and scientific optical cable routing planning is crucial for ensuring communication quality and improving communication service levels.

[0003] Existing fiber optic cable routing planning methods primarily rely on the experience and judgment of engineers. Engineers determine routes through manual comparison and selection. They need to comprehensively consider various factors, such as building structure and duct layout, to choose a relatively suitable route. However, this approach lacks systematic data processing and optimization analysis tools. When faced with complex building structures and diverse constraints, manual planning often struggles to fully consider all factors. Furthermore, regarding fiber optic cable capacity configuration, existing methods lack precise convergence calculation models and typically rely on empirical coefficients for estimation. In the comparison and evaluation of multiple route options, existing technologies lack a quantitative comprehensive evaluation system, making it difficult to achieve balanced optimization across multiple dimensions such as path length, laying difficulty, and security level.

[0004] Existing technologies have significant shortcomings. Routing planning methods relying on manual experience lack systematic data processing and optimization analysis tools, making it difficult to comprehensively consider multiple factors such as pipeline burial depth limitations, spacing requirements with other pipelines, and avoidance requirements in load-bearing areas. This can easily lead to suboptimal routing schemes or safety hazards. Furthermore, using empirical coefficients to estimate fiber optic cable capacity can result in either over- or under-allocation of fiber optic resources. Simultaneously, the lack of a quantitative comprehensive evaluation system makes it difficult to achieve balanced optimization across multiple dimensions when comparing various routing schemes. Summary of the Invention

[0005] To effectively improve the scientific nature of optical cable routing planning and the rationality of optical fiber resource allocation in triple-play projects, this application provides a method and system for optimizing optical cable routing planning in triple-play projects.

[0006] Firstly, this application provides a method for optimizing optical cable routing in a triple-play project, including: Constructing a network topology diagram; the construction process includes: extracting the coordinates of the communication equipment room from the building design data as the source node, extracting the coordinates of the fiber optic distribution boxes on each floor and the corresponding number of households covered as the target node and node attributes; vectorizing the outdoor pipes, vertical shafts and horizontal cable trays to obtain pipe attributes and establish path vector data; constructing a network topology diagram based on the source node, target node and path vector data; Complete the constraint quantification; the quantification process includes: extracting pipeline laying constraint parameters and computer room environment constraint parameters from engineering specification data; calculating the constraint weights of each edge in the network topology diagram based on the constraint parameters to obtain the comprehensive cost weights; Path optimization is performed; the optimization process includes: searching for candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights; evaluating the candidate paths in multiple dimensions to obtain normalized scores for each dimension; weighting and summing the normalized scores according to the weight coefficients of each dimension, and selecting the path with the highest comprehensive evaluation score as the final planned path from the source node to each target node. Optimize resource allocation; the allocation process includes: calculating the fiber core requirement layer by layer based on the final planned path and the number of households covered by the target node; selecting the capacity specification that meets the requirement from the distribution equipment specification sequence based on the fiber core requirement to form a resource allocation scheme.

[0007] By adopting the above scheme, the spatial data can be accurately described by vector modeling of the building communication nodes and pipeline paths. The constraints can be quantitatively represented by parametric processing required by engineering specifications. The route optimization solution under multiple constraints can be achieved by using a path search algorithm based on weighted graphs, thereby improving the efficiency and quality of route planning. At the same time, the precise allocation of fiber optic resources can be achieved by layer-by-layer convergence calculation, thereby reducing resource waste and costs.

[0008] Preferably, the step of calculating the constraint weights of each edge in the network topology graph based on the constraint parameters to obtain the comprehensive cost weights includes: The classification of each constraint parameter is determined, including hard constraints and soft constraints. For constraint parameters of different constraint types, the processing methods are determined separately. The processing methods for each constraint parameter of the hard constraint type include: determining whether the actual constraint parameter value meets the corresponding constraint parameter threshold requirement; if not, setting the additional constraint weight of the constraint parameter to infinity; if it meets the requirement, setting the additional constraint weight of the constraint parameter to zero. The processing methods for each constraint parameter of the soft constraint type include: calculating the difference between each actual constraint parameter value and the corresponding constraint parameter threshold, performing normalization processing and calculating the degree of violation; designing a single constraint penalty function; calculating the penalty value as the additional constraint weight; and when there are multiple constraint parameters on one side, weighting and calculating the total penalty value as the additional constraint weight. According to the determined processing method, determine the number and classification of constraint parameters for each edge, and calculate the additional constraint weights for each edge according to the determined processing method; combine the original length weights of each edge with the additional constraint weights to obtain the comprehensive cost weights.

[0009] By adopting the above scheme, the constraint parameters are classified into hard constraints and soft constraints. For different types, corresponding processing methods are used to calculate the additional weight of the constraints, which is then combined with the original length weight to obtain the comprehensive cost weight. This more accurately considers various constraint factors in the pipeline laying process, so as to make the path optimization results more reasonable.

[0010] Preferably, the step of using a graph search algorithm based on comprehensive cost weights to search for candidate paths from the source node to each target node includes: the graph search algorithm employs an improved Dijkstra algorithm, and searching for candidate paths using the improved Dijkstra algorithm includes: reading the source node's number from the network topology graph, setting the source node as the search starting point, initializing the source node's cumulative cost value to zero, and setting the cumulative cost values ​​of other nodes to maximum values; establishing a priority queue of nodes to be processed, processing nodes through the priority queue, updating the cumulative cost value of adjacent nodes and the predecessor node; and backtracking from the target node back to the source node to obtain candidate paths.

[0011] By adopting the above scheme, the improved Dijkstra algorithm is used to accurately and efficiently search for candidate paths from the source node to each target node in the network topology graph, thereby improving the accuracy and efficiency of path search.

[0012] Preferably, the construction process further includes: dividing the building into levels based on the floor scale in the building design data; for each level, segmenting the routing channels of the corresponding level according to the differences in business demand density and environmental constraints; and supplementing all nodes and routing segments with association attributes; the association attributes include: floor scale, business demand and environmental constraints. The quantification process also includes: based on the extracted pipeline laying constraint parameters and computer room environment constraint parameters, supplementing constraint parameters related to the associated attributes, including: floor scale constraint parameters, business requirement constraint parameters and environmental constraint supplementary parameters, and normalizing the constraint parameters related to the associated attributes. The multi-dimensional evaluation is set to evaluate the candidate path in three dimensions: path length, laying difficulty, and safety level. The weight coefficients of each dimension are dynamically obtained based on a neural network model. The input of the neural network model is the constraint parameters of the candidate path, and the output of the model is the optimal three-dimensional weight coefficients. The neural network model is generated by collecting building engineering cases with completed optical cable routing planning, extracting different candidate paths and their corresponding constraint parameters in each case, as well as the optimal three-dimensional weight coefficients verified by engineering projects, as training data.

[0013] By adopting the above scheme, the routing channels are layered and segmented, and associated attributes are added. Constraint parameters related to the associated attributes are added and normalized, so as to more finely consider the actual situation of different levels and road segments. Based on the actual situation, a neural network model is trained and used to dynamically obtain the weight coefficients of each dimension under the corresponding actual situation. The evaluation weights are dynamically adjusted according to different constraint parameters, so as to more accurately evaluate the candidate paths.

[0014] Preferably, the optimization process further includes: Perform topology segmentation to obtain hierarchical subtopologies and segmented subtopologies, and determine the associated subtopologies of each subtopology and the connecting nodes between the associated subtopologies; In the process of searching candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights, the sub-topology positioning of the source node and target node is determined, the sub-topology affiliation of the source node and target node is obtained, and it is determined whether cross-sub-topology calculation needs to be performed on the output. If it is determined that the sub-topology affiliation of the output source node and target node does not require cross-sub-topology calculation, the candidate path search is completed directly within the affiliation sub-topology using the graph search algorithm. If it is determined that the output source node and target node need to perform cross-sub-topology calculation, according to the path decomposition logic of decomposing the cross-sub-topology path into the source node, sub-topology connecting node, and target node, cross-sub-topology candidate connecting node pairs are screened. Based on the effective candidate connecting node pairs, the local optimal path result is obtained by combining the parallel computation search of each sub-topology, and all cross-sub-topology candidate paths are generated. For each cross-sub-topology candidate path, the scores of each segment of the path are integrated and combined with dynamic weights for comprehensive judgment, and finally the globally comprehensive optimal path is selected.

[0015] By adopting the above technical solution, the network topology map is divided into topological segments, which enables flexible path search processing by region and stage. Different path search strategies are adopted for source nodes and target nodes belonging to different sub-topologies, improving the targeting and efficiency of path search. By combining path decomposition logic with parallel computing of each sub-topology, the local optimal path result can be obtained quickly, and then all cross-sub-topology candidate paths can be generated and the globally comprehensive optimal path can be selected from them, enhancing the accuracy and comprehensiveness of the routing planning results.

[0016] Preferred options also include: Acquire building engineering project data of historical optical cable routing planning, and use deep learning algorithms to predict the future service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and equipment room constraint parameters. The construction process also includes: constructing the current topology version of the network topology map based on source nodes, target nodes, and path vector data; obtaining the future topology version by updating the current topology version based on the predicted future service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and data center constraint parameters; and determining the connecting nodes between the current topology version and the future topology version. The quantification process also includes: calculating the constraint weights of each edge in the network topology graph of the corresponding future topology version based on the predicted service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and data center constraint parameters for the future period, and obtaining the comprehensive cost weights under the corresponding version. The path optimization also includes: based on the predicted future service bandwidth growth rate, floor expansion or new node, and changes in pipeline and data center constraint parameters, using a neural network model to automatically adjust the weight coefficients of each dimension; based on different versions of the network topology map and their corresponding constraint quantification results, path optimization is performed for each time period to determine common path segments and differential path segments, and differential path segments are marked as elastic routing segments to reserve laying space and backup routes.

[0017] By adopting the above scheme, deep learning algorithms are used to predict changes in service bandwidth, floors, nodes, and constraint parameters in future periods, and optical cable routes that adapt to future development are planned in advance. By updating the topology version, recalculating weights, adjusting coefficients, and distinguishing path segments and reserved space, the forward-looking nature of route planning and the flexibility to respond to future changes are improved.

[0018] Preferably, the layer-by-layer aggregation calculation includes: grouping target nodes by building floor, accumulating the number of covered users of the same floor target node to obtain the total user access demand; calculating the fiber core demand of each floor according to the preset fiber core configuration standard per user; and superimposing the fiber core demand to the upper aggregation point layer by layer along the planned path from the bottom to the upper aggregation point to obtain the total fiber core demand at the communication equipment room exit. The resource allocation scheme includes: for source nodes in the communication equipment room, selecting the fiber distribution frame configuration capacity based on the total fiber core count requirement, and allocating the distribution frame capacity according to the ratio of operator area to user area; for target nodes on each floor, selecting the fiber distribution box configuration capacity based on the node's fiber core count requirement.

[0019] By adopting the above scheme, the fiber core requirements of each floor and communication equipment room exit can be accurately calculated based on the number of households covered by the target node and the routing path. The capacity of fiber distribution frames and fiber distribution boxes can be reasonably configured to achieve precise allocation of fiber resources and avoid over- or under-allocation of resources.

[0020] Secondly, this application provides a fiber optic cable routing optimization planning system for triple-play projects, comprising: The topology modeling module is used to construct a network topology diagram. The construction process includes: extracting the coordinates of the communication equipment room from the building design data as the source node; extracting the coordinates of the fiber optic distribution boxes on each floor and the corresponding coverage number attributes as the target node and node attributes; vectorizing the outdoor pipes, vertical shafts, and horizontal cable trays to obtain pipe attributes and establish path vector data; and constructing the network topology diagram based on the source node, target node, and path vector data. The constraint quantification module is used to quantify the constraints; the quantification process includes: extracting pipeline laying constraint parameters and computer room environment constraint parameters from engineering specification data; calculating the constraint weights of each edge in the network topology graph based on the constraint parameters to obtain the comprehensive cost weights; The routing calculation module is used for path optimization. The optimization process includes: searching for candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights; evaluating the candidate paths in multiple dimensions to obtain normalized scores for each dimension; weighting and summing the normalized scores according to the weight coefficients of each dimension, and selecting the path with the highest comprehensive evaluation score as the final planned path from the source node to each target node. The configuration optimization module is used to optimize resource allocation; the configuration process includes: calculating the fiber core requirement layer by layer based on the final planned path and the number of households covered by the target node; selecting the capacity specification that meets the requirement from the distribution equipment specification sequence based on the fiber core requirement to form a resource allocation scheme.

[0021] By adopting the above scheme, a topology modeling module is set up to vectorize the building communication nodes and pipeline paths, thereby achieving accurate description of spatial data; a constraint quantification module is set up to parametrically process the engineering specification requirements, thereby achieving quantitative expression of constraints; a routing calculation module is set up to use a path search algorithm based on a weighted graph to achieve route optimization solutions under multiple constraints; and a configuration optimization module is set up to achieve accurate configuration of fiber optic resources through layer-by-layer convergence calculation, effectively improving the scientific nature of fiber optic cable routing planning and the rationality of fiber optic resource allocation in triple-play projects.

[0022] Thirdly, this application provides a computer-readable storage medium including a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to perform the method described above.

[0023] Fourthly, this application provides a computer device, the computer device including a memory, a processor and a program stored in the memory and executable thereon, the program being executed by the processor to implement the steps of the method described above.

[0024] In summary, this application has the following beneficial effects: 1. Based on the vectorized modeling of building communication nodes and pipeline paths, the spatial data is accurately described, the engineering specifications are parameterized, the constraints are quantitatively expressed, and the route optimization solution under multiple constraints is performed based on the weighted graph path search algorithm. Through layer-by-layer convergence calculation, the precise allocation of optical fiber resources is achieved, thereby improving the scientific nature of optical cable route planning and the rationality of optical fiber resource allocation in the triple-play project. 2. By processing constraint parameters according to different constraint types, searching candidate paths using the Dijkstra algorithm, supplementing associated attributes and related constraint parameters, dynamically obtaining weight coefficients for each dimension, performing topology segmentation and cross-sub-topology calculations, predicting future changes and adjusting weighting coefficients, and reserving flexible routing segments, the accuracy and adaptability of the planning are further improved. Attached Figure Description

[0025] Figure 1 This is a flowchart of the optical cable route optimization planning method for the triple-play project described in a specific embodiment; Figure 2 This is a schematic diagram of the optical cable routing optimization planning system for the three-network integration project described in a specific embodiment. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0027] like Figure 1 As shown in the embodiment of this application, a method for optimizing and planning optical cable routes in a triple-play project is disclosed, including steps such as constructing a network topology map, quantifying constraints, performing path optimization, and optimizing resource allocation. This method achieves the effect of improving the scientific nature of optical cable route planning and the rationality of optical fiber resource allocation in triple-play projects. The following is a further detailed description of each step.

[0028] S1. Construct a network topology diagram.

[0029] Specifically, constructing the routing network topology serves as the data foundation for the entire method, including: extracting the spatial coordinates of communication nodes, vectorizing pipeline paths, and constructing the network topology graph.

[0030] In this embodiment, building engineering fiber optic cable routing planning is taken as an example. The extraction of communication node spatial coordinates refers to the spatial positioning and attribute labeling of communication nodes within the building. First, the coordinates of the communication equipment room are extracted as the source node of the routing network. The coordinates of the communication equipment room are usually obtained from the building design drawings, using the building's reference coordinate system to determine its planar and elevation coordinates. For example, the communication equipment room is generally located in the basement area of ​​the building, and its elevation coordinates are expressed as the vertical distance relative to the building's zero elevation. Second, the coordinates and coverage number of fiber optic distribution boxes on each floor are extracted as target nodes of the routing network. The coordinates of the fiber optic distribution boxes on each floor are extracted by scanning the design data of each floor of the building to determine their planar and elevation coordinates. Simultaneously, each target node is assigned a floor number attribute and a coverage number attribute, with the coverage number calculated based on the number of user access points within the service range of that fiber optic distribution box. The coordinate data of the source node and all target nodes are stored in a node coordinate data table. Each record in the data table includes a node number, node type, planar abscissa, planar ordinate, elevation coordinate, and auxiliary attribute fields.

[0031] The process involves vectorizing the communication ducts both inside and outside the building, converting their spatial orientation into calculable path data. The processed results serve as the foundation for selectable paths. The planar orientation of outdoor communication ducts is extracted from the duct design drawings. Continuous duct paths are segmented according to turning points, with each segment represented by its start and end coordinates, forming a duct path vector segment. Pipe attribute information, such as the number of pipe holes, pipe diameter, and laying method (e.g., underground laying), is added. The location and extent of vertical communication shafts are extracted from the building's internal design data. These shafts are abstracted as vertical vector segments, with the start elevation representing the bottom and the end elevation representing the top, and their cross-sectional dimensions and available capacity attributes are added. The orientation of horizontal cable trays and troughs is extracted from the floor design data, converted into a sequence of horizontal vector segments, and their specifications and remaining capacity attributes are added. All duct vectors, shaft vectors, and cable tray vectors are stored uniformly in a path vector data table.

[0032] The network topology graph is constructed based on the extracted node data and processed path data. Specifically, the extracted source and target nodes are used as the vertex set of the network topology graph, and the processed path vectors are used as the candidate edge set. Endpoint matching analysis is performed on each vector segment in the path vector data table to check whether the start and end points of the vector segment coincide with or are adjacent to the node positions in the node coordinate data table. When the spatial distance between a vector endpoint and a node is less than a preset matching tolerance, a connection is established between the vector and the node. For vector segments whose endpoints do not match any existing nodes, an intermediate transition node is created at its endpoint position, and the intermediate node is added to the vertex set. When two vector segments intersect or connect in space, a junction node is created at the intersection position, splitting the original vector segment into multiple sub-segments with the junction node as the boundary. All vector segments that have completed endpoint matching and intersection processing are used as edges of the network topology graph. Each edge records the vertex numbers of the two connected vertices and the edge length weight, with the edge length calculated based on the spatial distance between the vector segments.

[0033] In addition, to ensure the real-time nature of the constructed network topology map, data such as building design drawings, design data of each floor of the building, pipeline design drawings, building interior design data, and floor design data are collected and monitored in a timely manner. Once data updates occur, the extracted node data and path vector data are updated synchronously to assist in timely updating of the constructed network topology map.

[0034] S2. Complete the quantification of constraints.

[0035] Specifically, considering that the fiber optic cable routing planning in building engineering is constrained by engineering specification data, constraint parameters are extracted and quantified, and path edge constraint weights are calculated to assist in subsequent path optimization in the network topology graph.

[0036] First, the extraction of constraint parameters includes: extracting pipeline laying constraint parameters, computer room environment constraint parameters, and other constraint parameters from engineering specification data. Specifically, the engineering specification requirements for pipeline laying are parameterized, converting qualitative specification clauses into quantitative constraint thresholds. These pipeline laying constraint parameters include: the minimum required burial depth of communication pipelines read from the engineering specification database (stored in the pipeline burial depth lower limit parameter register); the critical burial depth value for which concrete encapsulation protection measures are required for the pipeline (stored in the encapsulation critical depth parameter register); the minimum horizontal and vertical clearance requirements between communication pipelines and gas pipelines (stored in the gas pipeline horizontal avoidance parameter register and gas pipeline vertical avoidance parameter register, respectively); the minimum clearance requirement between communication pipelines and high-voltage power cables (stored in the power cable avoidance parameter register); the length range for which concrete encapsulation is required on the outside of the flared opening when the pipeline enters a manhole or handhole (stored in the manhole encapsulation range parameter register); and reinforcement protection requirements for road crossing sections (stored in the road crossing reinforcement parameter register). All of these parameters are combined into a pipeline laying constraint parameter set.

[0037] The environmental and equipment installation requirements of the communication equipment room are parameterized. The environmental constraint parameters of the equipment room include: the minimum required value of the net height under the beam of the equipment room read from the equipment room design specification database (stored in the equipment room net height lower limit parameter register), the minimum required values ​​of the width and height of the fire door of the equipment room (stored in the door width lower limit parameter register and door height lower limit parameter register respectively), the minimum required value of the equivalent uniformly distributed live load of the general area of ​​the equipment room floor (stored in the general load-bearing lower limit parameter register), the minimum required value of the load-bearing capacity under the battery area (stored in the battery area load-bearing lower limit parameter register), the minimum required value of the horizontal illuminance of the general lighting in the equipment room (stored in the lighting illuminance lower limit parameter register), the minimum required value of the cross-sectional area of ​​the grounding conductor of the fiber optic distribution frame (stored in the grounding conductor cross-sectional parameter register), and the maximum allowable value of the grounding resistance of the equipotential grounding terminal box (stored in the grounding resistance upper limit parameter register), etc. All the above parameters are combined into a set of equipment room environmental constraint parameters.

[0038] Other constraint parameters may include: floor scale constraint parameters (number of floors, floor area, number of vertical pipes, etc.), business demand constraint parameters (business priority, total bandwidth demand, terminal distribution concentration, etc.), and environmental constraint supplementary parameters (construction difficulty, surrounding interference, etc.).

[0039] Secondly, constraint weights are calculated for each edge in the network topology graph based on constraint parameters to obtain the comprehensive cost weight. The calculation results are then updated in the edge attributes of the network topology graph. Finally, the edge set of the network topology graph is traversed, and the corresponding pipeline vector attribute information is extracted for each edge.

[0040] The constraint weight calculation specifically involves comparing the corresponding constraint parameter values ​​in actual building engineering with the standard constraint parameter thresholds, calculating the difference, and then applying additional weights to each constraint parameter based on engineering experience. For example, for an edge of an outdoor buried pipeline, the actual burial depth of the pipeline segment is read and compared with the lower limit parameter for pipeline burial depth. When the actual burial depth is less than the lower limit requirement, the insufficient burial depth is calculated, and the additional weight for encapsulation cost is obtained by consulting the encapsulation cost coefficient table based on the insufficient depth. When the actual burial depth is less than the critical encapsulation depth, an additional weight for encapsulation construction is added. Another example is reading the distribution data of other pipelines around the pipeline segment, calculating the minimum distance between the pipeline segment and gas pipelines and the minimum distance between the pipeline segment and power cables. The calculated distance values ​​are compared with the corresponding avoidance parameters. When the distance value is less than the avoidance parameter, it is determined that there is an avoidance constraint conflict on the edge, the availability flag of the edge is set to a restricted state, and a detour cost weight is added.

[0041] Therefore, the various constraint parameters are compared and verified, and the verification results of each constraint parameter are converted into constraint weight values ​​(if there are multiple constraint parameters on one side, they can be calculated and combined by weight). These values ​​are added to the original length weight of the edge to obtain the comprehensive cost weight value of the edge, and the edge weight attribute field stored in the network topology graph is updated.

[0042] S3. Perform path optimization to obtain the optimal planned path.

[0043] First, a candidate path search is performed. A graph search algorithm is used to search for candidate paths from the source node to each target node based on a comprehensive cost weight. In this embodiment, the graph search algorithm uses a modified Dijkstra's algorithm to complete the search for candidate paths from the source node to each target node.

[0044] The specific search steps include: Reading the source node's number from the network topology graph, setting the source node as the search starting point, initializing the cumulative cost value from the source node to itself to zero, and initializing the cumulative cost value of all other nodes to their maximum value. Establishing a priority queue of nodes to be processed, and adding the source node to the priority queue. Retrieving the node with the smallest cumulative cost value from the priority queue as the current processing node, traversing all adjacent edges of the current node, and for each adjacent edge, reading its comprehensive cost weight value and the adjacent node at the other end of the edge. Calculating the cumulative cost value from the current node to adjacent nodes, which is equal to the current node's cumulative cost value plus the weight value of the adjacent edge. When the calculated cumulative cost value is less than the currently recorded cumulative cost value of the adjacent node, updating the adjacent node's cumulative cost value and recording its predecessor node as the current node. Adding the updated adjacent node to the priority queue or updating its priority in the queue. Repeating the above process until the priority queue is empty or all target nodes have been processed. Based on the predecessor node information recorded by each target node, backtrack from the target node to the source node, record the node sequence and edge sequence on the backtracking path as the optimal path for that target node, and store them in the candidate path data table.

[0045] For example: Nodes: Communication equipment room A, floor distribution boxes B, C, and D; Calculated edge weights (comprehensive cost): A B: 10; A C: 5; B D: 3; C D: 8; B C: 2; The goal is to find the optimal path from A to D. Process node A, dequeue node A with the minimum cost of 0, and traverse its adjacent edges. B: New cost 0+10 Update B (cost = 10, predecessor = A) Join (B, 10); A C: New cost 0+5 Update C (cost = 5, predecessor = A) Enqueue (C, 5); process node C, the minimum cost of dequeueing is C; similarly, traverse adjacent edges, C D: New cost 5+8 Update D (cost = 13, predecessor = C) Join (D, 13); C B: New cost 5+2 10 Update B (cost = 7, predecessor = C) Update queue (B, 7); Process node B: Dequeue the node with the minimum cost (cost = 7), traverse adjacent edges, B D: New cost, new cost, 7+3 13 Update D (Cost = 10, Predecessor = B) Update queue (D, 10); Process node D: Dequeue node D (cost = 10), target node processed, terminate iteration. Path backtracking: Target node D, predecessor B, path segment: D B; Node B, predecessor C, path segment B C; Node C, predecessor A, path segment C A; Get the complete path.

[0046] To address the inability to handle dynamic changes (such as real-time weight updates) or the need for fault redundancy, a multi-predecessor dynamic K-value path generation method is chosen. This involves further modifying the Dijkstra algorithm to store multiple predecessor paths and their cumulative costs for each node, dynamically adjusting the number of candidates, k. The algorithm is further modified as follows: during initialization, a list is maintained for each node to store predecessor paths and their cumulative costs (instead of a single value); relaxation step: when the cost of a new path... Existing minimum cost + threshold (e.g., threshold) When the cost is 2), add it as a candidate predecessor; Priority queue: Each node can be enqueued multiple times (different paths), and the queue is sorted in ascending order of cost; Termination condition: Stop when k paths are output or the queue is empty (k can be dynamically set); Path backtracking: Traverse all candidate predecessor paths in reverse from the target node, and output the first k paths in ascending order of cost. Therefore, the result of the above example is: Optimal path: A C B D (cost 10), secondary path: A C D (cost 13).

[0047] Secondly, the candidate paths are evaluated from multiple dimensions to obtain normalized scores for each dimension. In this embodiment, the multi-dimensional evaluation is set to a three-dimensional evaluation of the candidate paths: path length, laying difficulty, and safety level. Specifically, the node sequence and edge sequence of each candidate path are read sequentially from the candidate path data table. For the path length dimension, the original length values ​​of all edges contained in the path are summed to obtain the total physical length of the path. The total physical length value is normalized, using the maximum length among all candidate paths as the normalization benchmark, and the path's length normalization score is calculated; the shorter the length, the higher the score. For the laying difficulty dimension, the number of edges in the path that are laid as underground pipelines, overhead cable trays, and vertical shafts is counted. The comprehensive laying difficulty value is calculated by weighting the construction difficulty coefficients of each laying method, and then normalized to obtain the laying difficulty score. For the safety level dimension, the number of edges in the path containing constraint conflict markers and the cumulative sum of constraint weight values ​​are counted. The comprehensive safety risk value is calculated based on the severity of the constraint conflict, and then normalized to obtain a safety level score. The scores for each dimension are stored in the path evaluation result data table.

[0048] Finally, the normalized scores are weighted and synthesized according to the weight coefficients of each dimension, and the path with the highest comprehensive evaluation score is selected as the final planned path from the source node to each target node. Specifically, the weight coefficients for path length, laying difficulty, and security level are read from the preset configuration parameters, and the sum of the three weight coefficients equals the unit value. The normalized scores of each candidate path are read from the path evaluation result data table, and the length dimension score is multiplied by the length weight coefficient, the laying difficulty score by the difficulty weight coefficient, and the security level score by the security weight coefficient. The three products are added together to obtain the comprehensive evaluation score of the path. When there are multiple candidate paths for the same target node, the comprehensive evaluation scores of each candidate path are compared, and the path with the highest score is selected as the final planned path for that target node. The node sequence, edge sequence, physical length, and comprehensive score of the final planned path are stored in the final routing scheme data table.

[0049] S4. Optimize resource allocation and complete patch panel capacity configuration.

[0050] Specifically, the final route plan data table reads all the final planned paths, the node coordinate data table reads the coverage number of each target node, the fiber optic capacity is calculated layer by layer, and the patch panel capacity configuration is completed.

[0051] The layer-by-layer aggregation calculation includes: grouping target nodes by building floor, grouping target nodes on the same floor together; summing the number of users covered by each target node in the group to obtain the total user access demand for that floor; multiplying the total user access demand by the fiber core count configuration standard per user to calculate the fiber core count demand for the floor; performing convergence aggregation calculation layer by layer from the bottom floor upwards; when a target node on a certain floor is connected to the upper-level aggregation point through a vertical shaft, the fiber core count demand for that floor is added to the convergence demand of the upper-level aggregation point; and aggregating the convergence demand of each floor towards the source node according to the aggregation structure of the routing path, ultimately obtaining the total fiber core count demand at the communication equipment room exit.

[0052] The configuration process includes: retrieving a sequence of available fiber optic distribution frame capacity specifications from the distribution equipment specification database, with the sequence arranged in ascending order of capacity. For source nodes within the communication equipment room, the minimum specification value not less than the required capacity is found in the capacity specification sequence, and this value is used as the configuration capacity of the fiber optic distribution frame in the equipment room. Considering the service division requirements of triple-play integration, the distribution frame capacity is divided according to a preset ratio of operator area and user area, resulting in fiber optic distribution frame capacities for the operator area and user area respectively. For target nodes on each floor, the minimum specification value not less than the required capacity is found in the capacity specification sequence, and this value is used as the configuration capacity of the fiber optic distribution box on that floor. The configuration capacities of the distribution equipment for all nodes are summarized to form a resource configuration list, which is then associated with the aforementioned final routing scheme data table to output a complete triple-play integration optical cable routing optimization planning scheme.

[0053] In a specific embodiment, to further improve the accuracy and rationality of path optimization, constraint parameters are classified and processed differently for hard and soft constraints. This allows for reasonable penalties for non-compliance of soft constraints while meeting basic constraint requirements, and more accurately calculates the constraint weights of edges. The method further includes: besides calculating the difference by uniformly comparing the corresponding constraint parameter thresholds and applying additional constraint weights to each constraint parameter based on engineering experience, the calculation of constraint weights for each edge in the network topology graph based on the constraint parameters to obtain the comprehensive cost weights also includes: first, determining the classification of each constraint parameter, including hard and soft constraints. Hard constraints are those that must be strictly met; otherwise, the edge is not passable. For example, the bending radius of a pipeline; comparing the actual bending radius with the minimum allowable bending radius will lead to fiber optic breakage. Soft constraints, on the other hand, allow a certain degree of violation, but will increase construction difficulty, risk, or maintenance costs. They need to be quantified and added as additional constraint weights, such as the burial depth of a pipeline.

[0054] Therefore, processing methods are determined for constraint parameters of different constraint types. For hard constraint types, the processing method for each constraint parameter includes: determining whether the actual constraint parameter value meets the corresponding constraint parameter threshold requirement; if not, setting the additional weight of the constraint parameter to infinity; if it meets the requirement, setting the additional weight of the constraint parameter to zero. For soft constraint types, the processing method includes: calculating the difference between each actual constraint parameter value and the corresponding constraint parameter threshold, performing normalization processing and calculating the degree of violation; designing a single-constraint penalty function; calculating the penalty value as the additional constraint weight; and when there are multiple constraint parameters on one side, weighted calculating the total penalty value as the additional constraint weight.

[0055] In this embodiment, the dimensions and numerical ranges of different constraints vary greatly, so normalization is required first. The degree of violation of the i-th soft constraint is defined. For lower bound constraints (actual values) It should be no less than the lower limit ), ,in, As a normalization benchmark, the typical range of variation or the maximum allowable deviation of this parameter can be taken. This indicates a serious violation. For upper bound constraints (actual value) It should not exceed the upper limit. ), For interval constraints (actual value) Should be in [ ]), Regarding the degree of violation after normalization Design a penalty function A nonlinear function can be used as the penalty function. In this embodiment, a piecewise linear function is used, and the formula is: In the formula, This is the tolerance threshold for the constraint parameter; anything below this value is considered a minor violation and will be subject to a lighter penalty. The slope of the linear region. , The rate at which penalties increase beyond the tolerance zone.

[0056] Secondly, following the established processing method, the number and classification of constraint parameters for each edge are determined, and the additional constraint weights for each edge are calculated according to the established processing method. The original length weights of each edge are then combined with the additional constraint weights to obtain the comprehensive cost weights. Specifically, in addition to weighted merging, the weights can also be uniformly converted into cost dimensions.

[0057] One specific implementation involves segmenting and layering floors, supplementing them with associated attributes, to more meticulously consider the characteristics of different floors and routing segments. This includes supplementing and normalizing constraint parameters related to associated attributes, dynamically obtaining weight coefficients for each dimension using a neural network model, and performing topology optimization, thereby improving the scientific rigor and adaptability of routing planning. The method further includes incorporating "dynamic adjustment of weighted comprehensive coefficients" and "layered and segmented routing optimization" by combining floor size, business requirements, and environmental constraints. This makes fiber optic routing planning more flexible, practical, and accurate, enabling scenario-based matching of the planning process. The following sections elaborate on these aspects: routing network topology optimization, constraint quantification upgrade, dynamic adjustment of weighted comprehensive coefficients, and layered and segmented routing optimization.

[0058] First, optimize the routing network topology. Based on the existing communication equipment room + floor fiber distribution box topology model, and combined with the floor scale and service requirements, add layered and segmented node design to build a three-level network of "global topology - layered topology - segmented topology"; the global topology still takes the communication equipment room as the core (source) node and the fiber distribution boxes on each floor as the first-level (target) nodes.

[0059] The construction process also includes: dividing the building into levels based on the floor scale in the architectural design data, and splitting them using a scale threshold method; for example: small-scale floors ( For a medium-sized layer, it is divided into two layers (low zone: layers 1-8, high zone: layers 9-15), with an additional partitioned fiber optic distribution box on each layer, designated as a secondary (target) node. The partitioned boxes in the low and high zones are connected to the main distribution box on their respective floors. For a large-scale layer, it is divided into three layers (low zone: layers 1-8, medium zone: layers 9-16, high zone: layers 17 and above), with an additional partitioned fiber optic distribution box on each layer, designated as a secondary (target) node. The partitioned boxes are connected to the main distribution box on the corresponding floor, and the main distribution box is connected to the communication equipment room.

[0060] The construction process also includes: in addition to the original segmentation based on inflection points and junction nodes, for each level, the routing channels of the corresponding level are segmented according to the differences in business demand density and environmental constraints. For example, each segment is assigned a monitoring node (a third-level node), and the specific attributes of each segment are labeled (such as segment length, deployment method, environmental interference sources, and number of business terminals). Associated attributes are added to all nodes and routing segments, including: floor level, business needs, and environmental constraints.

[0061] Secondly, improve the quantification of constraint parameters. Supplement constraint parameters related to floor scale and business requirements, quantify them into calculable values, and provide a quantitative basis for dynamic adjustment of weighting coefficients. The quantification process also includes: based on the extracted pipeline laying constraint parameters and computer room environment constraint parameters, supplementing constraint parameters related to associated attributes, including: floor scale constraint parameters, business requirement constraint parameters, and supplementary environmental constraint parameters, and normalizing the constraint parameters related to associated attributes. For example, floor scale constraint parameters include: number of floors (quantified as 1-3, corresponding to small, medium, and large scale), floor building area (quantified as 1-N, corresponding to area ranges A, B, and C), and number of vertical pipes (1-N, corresponding to the number of pipes X, Y, and Z); business demand constraint parameters include: business priority (core business takes 3, important business takes 2, and ordinary business takes 1), total bandwidth demand (≤1000Mbps takes 1, 1000-5000Mbps takes 2, and ≥5000Mbps takes 3), and terminal distribution concentration (centralized takes 1, decentralized takes 3, and mixed takes 2); environmental constraint supplementary parameters include: floor construction difficulty (new building takes 1, old building takes 3, and renovated building takes 2), surrounding interference sources (strong interference takes 3, moderate interference takes 2, and weak interference takes 1), and laying space restrictions (unrestricted takes 1, restricted takes 2, and severely restricted takes 3). After normalization, the corresponding supplementary constraint parameters can be quantified using the constraint supplementary weight calculation method in the above embodiment to obtain the constraint supplementary weight of the corresponding parameter.

[0062] Then, the weighted comprehensive coefficients are dynamically adjusted. The weight coefficients for each dimension in the multi-dimensional evaluation setting are dynamically obtained based on a neural network model, replacing the direct reading of the weight coefficients for path length, laying difficulty, and safety level dimensions from preset configuration parameters. The input to the neural network model is the constraint parameters of the candidate path (including floor size, business requirements, and environmental constraints), and the output is the optimal three-dimensional weight coefficients (path length, laying difficulty, and safety level). Specifically, the neural network model is generated by collecting building engineering cases with completed fiber optic cable routing plans, extracting different candidate paths and their corresponding constraint parameters from each case, as well as the optimal three-dimensional weight coefficients verified by engineering projects, as training data. Collecting building engineering cases with completed fiber optic cable routing plans can also be done by using an 80%:20% ratio of actual building engineering cases with completed fiber optic cable routing plans and simulated building engineering cases with completed fiber optic cable routing plans for training.

[0063] Finally, hierarchical and segmented route optimization is performed. Considering the large scale of building projects and the large number of target nodes, optimizing each node individually would be time-consuming. Therefore, a hierarchical and segmented optimization approach is designed. Specifically, the optimization process also includes: performing topology segmentation, obtaining hierarchical sub-topologies and segmented sub-topologies, and determining the associated sub-topologies of each sub-topology and the connecting nodes between them. For example, the connecting nodes of hierarchical sub-topologies are the floor master allocation boxes, and the connecting nodes of segmented sub-topologies are the zone allocation boxes.

[0064] In the process of searching candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights, the sub-topology location of the source node and the target node is determined, that is, the sub-topology type and number to which they belong are clarified. The sub-topology affiliation of the source node and the target node is obtained, and it is determined whether cross-sub-topology calculation is required. For example: if the matching result of the start point and the end point is the same layer sub-topology or the same segment sub-topology, then cross-sub-topology calculation is not required; if the matching result of the start point and the end point is "different segment sub-topologies (belonging to the same layer sub-topology)," then "cross-segment sub-topology calculation" is required; or if the matching result of the start point and the end point is "different layer sub-topologies" (regardless of whether they are in the same segment), then "cross-layer sub-topology calculation" is required. That is, the output includes the sub-topology (number, type) to which the start point belongs, the sub-topology (number, type) to which the end point belongs, whether cross-sub-topology is required, the type of cross-sub-topology (cross segment / cross layer), and the range of candidate connecting nodes.

[0065] In this case, if it is not necessary to perform cross-subtopology calculation to determine the subtopology affiliation of the output source node and the target node, the candidate path search can be completed directly within the affiliated subtopology using a graph search algorithm.

[0066] If it is determined that cross-subtopology calculations are required between the source node and the target node, the path decomposition logic is followed, breaking down the cross-subtopology path into source node, subtopology connecting node, and target node. For example, when the starting point S and the ending point T are in different subtopologies, the topology path is decomposed into the starting point S. Subtopology A Connecting Nodes Subtopology B connecting nodes The endpoint is T. Candidate connection node pairs across sub-topologies are selected through a process including: First, determining the scope of connection nodes: Across segmented sub-topologies (within the same layer): all exit connection nodes of the starting segmented sub-topology (A) and all entrance connection nodes of the ending segmented sub-topology (B) constitute the candidate connection node pool; Across layered sub-topologies (different layers): all exit connection nodes of the starting layered sub-topology (A) and all entrance connection nodes of the ending layered sub-topology (B) constitute the candidate connection node pool; and connection nodes must meet the constraint conditions, with those not meeting the constraints being removed. Second, generating candidate connection node pairs: one node (denoted as C) is selected from the exit connection nodes of the starting sub-topology (A), and one node (denoted as D) is selected from the entrance connection nodes of the ending sub-topology (B), forming a connection node pair (C, D). All node pairs meeting the conditions are then selected to form a set of candidate node pairs. Third, preliminary screening of candidate node pairs is performed. The feasibility of a direct connection path between each candidate node pair (C,D) is calculated, i.e., whether there is a channel (such as a vertical pipe or a horizontal main pipe) between C and D where optical cables can be laid. If there is no direct connection channel, the node pair is directly eliminated. If there is a direct connection channel, the basic edge weight of the channel is calculated (based on the dynamic weight coefficient of the current cross-sub-topology region). Node pairs with a basic edge weight greater than 1 are eliminated, and the set of valid candidate node pairs is retained.

[0067] Based on valid candidate connection node pairs, and combined with the local optimal path results obtained through parallel computation search of each sub-topology, all cross-sub-topology candidate paths are generated. For example: each cross-sub-topology candidate path = the first segment (S C) + Second paragraph (C) D) + Third paragraph (D) T), each optimal path is optimized in parallel and independently. For each cross-sub-topology candidate path, the scores of each path segment are integrated and combined with dynamic weights for a comprehensive judgment (based on engineering experience, this can be flexibly adjusted). =0.3 (Influence of starting point sub-topology weight) =0.3 (Influence of terminal sub-topology weight) =0.4 (influence of cross-sub-topology channel weights), calculate the weighted score of each cross-sub-topology candidate path, and finally select the globally optimal path.

[0068] Furthermore, if no valid candidate connecting node pairs are found after screening (e.g., no available pipeline channels between two sub-topologies), the topology segmentation rules are readjusted (e.g., merging adjacent segmented sub-topologies), or connecting nodes are added (e.g., adding pipeline interfaces), and screening and path calculation are repeated. If the start and end points span multiple layers or segments, they are broken down into multiple cross-sub-topology calculations (e.g., lower-level segment 1). Low-level segment 2 Middle section 1 (3) The middle section is divided into segments, and the optimal path across each sub-topology is selected step by step, and finally integrated into the optimal path for the whole process.

[0069] One specific embodiment considers a method that addresses the inability of current routing planning to adapt to future business growth, topology changes, and constraint adjustments. Through a closed-loop design involving future demand forecasting, dynamic topology adaptation, dynamic route optimization, and flexible resource allocation, the method achieves adaptive adjustment of the routing scheme. The method further includes: First, by combining the initial design's topology nodes (communication equipment rooms, floor fiber distribution boxes), constraint parameters (duct laying, equipment room environment), and resource allocation (fiber capacity), the prediction targets are determined, including: service demand (bandwidth demand growth), topology prediction (addition / adjustment of fiber distribution boxes and communication equipment rooms, and expansion requirements of duct channels), and changes in constraints (changes in duct and equipment room constraint parameters). Specifically, historical building engineering project data on fiber optic cable routing planning can be obtained (including historical data (such as bandwidth growth over the past 1-2 years), building plans, and business development plans). Deep learning algorithms are then used to predict future time periods (1 year, 3 years, 5 years) for service bandwidth growth rates, floor expansion or addition of nodes, and changes in duct and equipment room constraint parameters.

[0070] Secondly, based on future predictions, dynamic topology adaptation, dynamic route optimization, and flexible resource configuration are designed. Specifically, the construction process includes: constructing the current topology version of the network topology map based on source nodes, target nodes, and path vector data; updating the current topology version to obtain the future topology version based on predicted future service bandwidth growth rates, floor expansion or new node additions, and changes in pipeline and equipment room constraint parameters; and determining the connecting nodes (such as expandable fiber distribution boxes) between the current and future topology versions. The quantization process also includes: calculating the constraint weights of each edge in the network topology map of the corresponding future topology version based on predicted future service bandwidth growth rates, floor expansion or new node additions, and changes in pipeline and equipment room constraint parameters to obtain the comprehensive cost weight under the corresponding version. The path optimization also includes: automatically adjusting the weight coefficients of each dimension using a neural network model based on predicted future service bandwidth growth rates, floor expansion or new node additions, and changes in pipeline and equipment room constraint parameters; performing path optimization for each time period based on the quantization results of different versions of the network topology map and their corresponding constraints, determining common path segments and differential path segments, marking differential path segments as flexible routing segments, and reserving deployment space and backup routes. The resource allocation also includes: Current configuration: Based on the current optimal path planning, fiber optic capacity is configured at 1.05 times the current service demand. Future period: In elastic routing segments and cross-sub-topology connection nodes, 20%-30% of the fiber cores are reserved according to the predicted future demand gap to avoid rewiring during future capacity expansion. For example: If the service bandwidth growth rate of a certain floor reaches 40%, the reserved fiber cores will be allocated to the routing segment of that floor to temporarily expand the fiber optic capacity.

[0071] like Figure 2 As shown in the figure, this application discloses a fiber optic cable routing optimization planning system for a triple-play project, specifically including: The topology modeling module 100 is used to construct a network topology diagram. The construction process includes: extracting the coordinates of the communication equipment room from the building design data as the source node; extracting the coordinates of the fiber optic distribution boxes on each floor and the corresponding coverage number attributes as the target node and node attributes; vectorizing the outdoor pipes, vertical shafts, and horizontal cable trays to obtain pipe attributes and establish path vector data; and constructing a network topology diagram based on the source node, target node, and path vector data. The constraint quantification module 200 is used to complete the quantification of constraint conditions; the quantification process includes: extracting pipeline laying constraint parameters and computer room environment constraint parameters from engineering specification data; calculating the constraint weights of each edge in the network topology diagram according to the constraint parameters to obtain the comprehensive cost weight; The routing calculation module 300 is used for path optimization. The optimization process includes: searching for candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights; evaluating the candidate paths in multiple dimensions to obtain normalized scores for each dimension; weighting and summing the normalized scores according to the weight coefficients of each dimension, and selecting the path with the highest comprehensive evaluation score as the final planned path from the source node to each target node. The configuration optimization module 400 is used to optimize resource configuration. The configuration process includes: calculating the fiber core requirement layer by layer based on the final planned path and the number of households covered by the target node; selecting the capacity specification that meets the requirement from the distribution equipment specification sequence based on the fiber core requirement to form a resource configuration scheme.

[0072] The system also includes a demand forecasting module 500, which is used to acquire building engineering project data of historical engineering optical cable routing planning, and use deep learning algorithms to predict the service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and equipment room constraint parameters in future periods.

[0073] This application also discloses a computer-readable storage medium.

[0074] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed, such as the optical cable routing optimization planning method for the three-network convergence project described above. The computer-readable storage medium includes, for example, various media that can store program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0075] This application also discloses a computer device.

[0076] Specifically, the computer device includes a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed to implement the aforementioned optical cable route optimization planning method for the three-network integration project.

[0077] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A method for optimizing optical cable routing in a triple-play project, characterized in that, include: Construct a network topology diagram; The construction process includes: extracting the coordinates of the communication equipment room from the building design data as the source node; extracting the coordinates of the fiber optic distribution boxes on each floor and the corresponding number of households covered as the target node and node attributes; vectorizing the outdoor pipes, vertical shafts and horizontal cable trays to obtain pipe attributes and establish path vector data; and constructing a network topology based on the source node, target node and path vector data. Complete the constraint quantification; the quantification process includes: extracting pipeline laying constraint parameters and computer room environment constraint parameters from engineering specification data; calculating the constraint weights of each edge in the network topology diagram based on the constraint parameters to obtain the comprehensive cost weights; Path optimization is performed; the optimization process includes: searching for candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights; evaluating the candidate paths in multiple dimensions to obtain normalized scores for each dimension; weighting and summing the normalized scores according to the weight coefficients of each dimension, and selecting the path with the highest comprehensive evaluation score as the final planned path from the source node to each target node. Optimize resource allocation; the allocation process includes: calculating the fiber core requirement layer by layer based on the final planned path and the number of households covered by the target node; selecting the capacity specification that meets the requirement from the distribution equipment specification sequence based on the fiber core requirement to form a resource allocation scheme.

2. The optical cable route optimization planning method for the triple-play project according to claim 1, characterized in that, The step of calculating the constraint weights of each edge in the network topology graph based on constraint parameters to obtain the comprehensive cost weights includes: The classification of each constraint parameter is determined, including hard constraints and soft constraints. For constraint parameters of different constraint types, the processing methods are determined separately. The processing methods for each constraint parameter of the hard constraint type include: determining whether the actual constraint parameter value meets the corresponding constraint parameter threshold requirement; if not, setting the additional constraint weight of the constraint parameter to infinity; if it meets the requirement, setting the additional constraint weight of the constraint parameter to zero. The processing methods for each constraint parameter of the soft constraint type include: calculating the difference between each actual constraint parameter value and the corresponding constraint parameter threshold, performing normalization processing and calculating the degree of violation; designing a single constraint penalty function; calculating the penalty value as the additional constraint weight; and when there are multiple constraint parameters on one side, weighting and calculating the total penalty value as the additional constraint weight. According to the determined processing method, determine the number and classification of constraint parameters for each edge, and calculate the additional constraint weights for each edge according to the determined processing method; combine the original length weights of each edge with the additional constraint weights to obtain the comprehensive cost weights.

3. The optical cable route optimization planning method for the triple-play project according to claim 1, characterized in that, The method of searching candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights includes: the graph search algorithm adopts an improved Dijkstra algorithm, and searching candidate paths using the improved Dijkstra algorithm includes: reading the source node number from the network topology graph, setting the source node as the search starting point, initializing the cumulative cost value of the source node to zero, and setting the cumulative cost values ​​of other nodes to maximum values; establishing a priority queue of nodes to be processed, processing nodes through the priority queue, updating the cumulative cost value of adjacent nodes and the predecessor node; and backtracking from the target node to the source node to obtain candidate paths.

4. The optical cable route optimization planning method for the triple-play project according to claim 1, characterized in that, The construction process also includes: dividing the building into levels based on the floor scale in the building design data; for each level, segmenting the routing channels of the corresponding level according to the differences in business demand density and environmental constraints; and supplementing all nodes and routing segments with association attributes; the association attributes include: floor scale, business demand, and environmental constraints. The quantification process also includes: based on the extracted pipeline laying constraint parameters and computer room environment constraint parameters, supplementing constraint parameters related to the associated attributes, including: floor scale constraint parameters, business requirement constraint parameters and environmental constraint supplementary parameters, and normalizing the constraint parameters related to the associated attributes. The multi-dimensional evaluation is set to evaluate the candidate path in three dimensions: path length, laying difficulty, and safety level. The weight coefficients of each dimension are dynamically obtained based on a neural network model. The input of the neural network model is the constraint parameters of the candidate path, and the output of the model is the optimal three-dimensional weight coefficients. The neural network model is generated by collecting building engineering cases with completed optical cable routing planning, extracting different candidate paths and their corresponding constraint parameters in each case, as well as the optimal three-dimensional weight coefficients verified by engineering projects, as training data.

5. The optical cable route optimization planning method for the triple-play project according to claim 4, characterized in that, The optimization process also includes: Perform topology segmentation to obtain hierarchical subtopologies and segmented subtopologies, and determine the associated subtopologies of each subtopology and the connecting nodes between the associated subtopologies; In the process of searching candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights, the sub-topology positioning of the source node and target node is determined, the sub-topology affiliation of the source node and target node is obtained, and it is determined whether cross-sub-topology calculation needs to be performed on the output. If it is determined that the sub-topology affiliation of the output source node and target node does not require cross-sub-topology calculation, the candidate path search is completed directly within the affiliation sub-topology using the graph search algorithm. If it is determined that the output source node and target node need to perform cross-sub-topology calculation, according to the path decomposition logic of decomposing the cross-sub-topology path into the source node, sub-topology connecting node, and target node, cross-sub-topology candidate connecting node pairs are screened. Based on the effective candidate connecting node pairs, the local optimal path result is obtained by combining the parallel computation search of each sub-topology, and all cross-sub-topology candidate paths are generated. For each cross-sub-topology candidate path, the scores of each segment of the path are integrated and combined with dynamic weights for comprehensive judgment, and finally the globally comprehensive optimal path is selected.

6. The optical cable route optimization planning method for the triple-play project according to claim 4, characterized in that, Also includes: Acquire building engineering project data of historical optical cable routing planning, and use deep learning algorithms to predict the future service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and equipment room constraint parameters. The construction process also includes: constructing the current topology version of the network topology map based on source nodes, target nodes, and path vector data; obtaining the future topology version by updating the current topology version based on the predicted future service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and data center constraint parameters; and determining the connecting nodes between the current topology version and the future topology version. The quantification process also includes: calculating the constraint weights of each edge in the network topology graph of the corresponding future topology version based on the predicted service bandwidth growth rate, floor expansion or addition of nodes, and changes in pipeline and data center constraint parameters for the future period, and obtaining the comprehensive cost weights under the corresponding version. The path optimization also includes: based on the predicted future service bandwidth growth rate, floor expansion or new node, and changes in pipeline and data center constraint parameters, using a neural network model to automatically adjust the weight coefficients of each dimension; based on different versions of the network topology map and their corresponding constraint quantification results, path optimization is performed for each time period to determine common path segments and differential path segments, and differential path segments are marked as elastic routing segments to reserve laying space and backup routes.

7. The method for optimizing optical cable routing in a triple-play project according to claim 1, characterized in that, The layer-by-layer aggregation calculation includes: grouping target nodes by building floor, accumulating the number of covered households of the target nodes on the same floor to obtain the total user access demand; calculating the fiber core demand of each floor according to the preset fiber core configuration standard per user; and superimposing the fiber core demand to the upper aggregation point layer by layer along the planned path from the bottom to the upper aggregation point to obtain the total fiber core demand at the communication equipment room exit. The resource allocation scheme includes: for source nodes in the communication equipment room, selecting the fiber distribution frame configuration capacity based on the total fiber core count requirement, and allocating the distribution frame capacity according to the ratio of operator area to user area; for target nodes on each floor, selecting the fiber distribution box configuration capacity based on the node's fiber core count requirement.

8. A fiber optic cable routing optimization planning system for a triple-play project, characterized in that, include: The topology modeling module is used to construct network topology graphs; The construction process includes: extracting the coordinates of the communication equipment room from the building design data as the source node; extracting the coordinates of the fiber optic distribution boxes on each floor and the corresponding number of households covered as the target node and node attributes; vectorizing the outdoor pipes, vertical shafts and horizontal cable trays to obtain pipe attributes and establish path vector data; and constructing a network topology based on the source node, target node and path vector data. The constraint quantification module is used to quantify the constraints; the quantification process includes: extracting pipeline laying constraint parameters and computer room environment constraint parameters from engineering specification data; calculating the constraint weights of each edge in the network topology graph based on the constraint parameters to obtain the comprehensive cost weights; The routing calculation module is used for path optimization. The optimization process includes: searching for candidate paths from the source node to each target node using a graph search algorithm based on comprehensive cost weights; evaluating the candidate paths in multiple dimensions to obtain normalized scores for each dimension; weighting and summing the normalized scores according to the weight coefficients of each dimension, and selecting the path with the highest comprehensive evaluation score as the final planned path from the source node to each target node. The configuration optimization module is used to optimize resource allocation; the configuration process includes: calculating the fiber core requirement layer by layer based on the final planned path and the number of households covered by the target node; selecting the capacity specification that meets the requirement from the distribution equipment specification sequence based on the fiber core requirement to form a resource allocation scheme.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the method as described in any one of claims 1 to 7.

10. A computer device, characterized in that, The computer device includes a memory, a processor, and a program stored in and executable on the memory, the program being executed by the processor to implement the steps of the method as described in any one of claims 1 to 7.

Images