Transformer substation insulation tubular bus path trend optimization method

Abstract

The invention provides a transformer station insulation tubular bus path trend optimization method comprising the following steps: parallelly constructing an insulation tubular bus topology mapping frame and a bus-level collaborative optimization frame: the insulation tubular bus topology mapping frame splits a bus into bus segment calculation units capable of synchronous optimization, and each unit generates a discrete node chain through iteration according to a starting coordinate under comprehensive constraints; quantizing by a path evaluation model to obtain cost values, and converging the cost values into total topological cost; the latter issues constraints and coordinates through a constraint injection module, and a decision module outputs the total cost. When the framework is constructed, a path evaluation model is firstly established in a bus segment equivalent network, a node coordinate distributor and a cost value collector are configured, and then redundant nodes are identified and removed. According to the method, the global optimality and adaptability of path planning can be improved, the construction cost and the operation and maintenance difficulty are reduced, and the safety and reliability of a power system are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system substation design and optimization, and more particularly, relates to a substation insulated pipe-type busbar path orientation optimization method. BACKGROUND

[0002] In the power system, the insulated pipe-type busbar of the substation as a key power transmission component, the optimization of its path orientation is of great significance to ensure the stable operation of the power system, reduce the construction cost and improve the operation and maintenance efficiency. The traditional substation insulated pipe-type busbar path planning method mainly relies on manual experience, which is designed by engineers according to the civil structure, equipment layout and electrical safety requirements of the substation. Although this method can meet the basic engineering requirements to some extent, it has many limitations. First, the manual planning process is time-consuming, especially when facing complex substation layout, it needs to adjust and verify the path orientation repeatedly to ensure that all constraint conditions are met. Secondly, manual planning cannot guarantee the global optimality of the path, engineers can only design based on local experience, which may lead to path redundancy, material waste and increased construction difficulty. In addition, with the continuous expansion of the substation scale and the increase of the number of equipment, the operability and accuracy of the manual planning method are seriously challenged, which is difficult to meet the needs of modern power systems for efficient and accurate planning.

[0003] In terms of technical principles, the existing substation insulated pipe-type busbar path planning method is mainly based on simple geometric calculation and electrical safety specifications. These methods usually only consider the straight line segment and a few fixed bending points of the busbar, ignoring the influence of various complex factors that the busbar may be subjected to in actual operation, such as the stress of the supporting insulator, the displacement of the expansion joint and the state of the grounding switch. Although this simplified model simplifies the calculation process to some extent, it cannot accurately reflect the actual operating state of the busbar, resulting in a large deviation between the planning results and the actual situation. At the same time, the existing methods often use a fixed minimum bending radius value when dealing with the bending radius constraint of the busbar path, without considering the differences under different busbar models and actual working conditions, which may lead to stress concentration, fatigue fracture and other problems in the actual installation and operation of the busbar, affecting the safety and reliability of the power system.

[0004] In the implementation process of the embodiment of the present application, at least the following problems or defects exist in the prior art: First, the traditional method relies on manual experience, and the planning efficiency is low and it is difficult to guarantee the global optimality; second, the factors considered in the path planning of the prior art are relatively single, and the actual running state of the busbar cannot be fully reflected; finally, the existing path planning method has limitations in handling complex constraint conditions, and cannot effectively solve the multi-constraint balance problem in busbar path optimization. These problems limit the further development of the substation insulating tubular busbar path optimization, and there is an urgent need for a more efficient, accurate and adaptable optimization method for complex working conditions. SUMMARY

[0005] The present application provides a substation insulating tubular busbar path optimization method, comprising: A parallel insulating tubular busbar topology mapping framework of the target substation is built; the insulating tubular busbar topology mapping framework comprises a plurality of busbar segment calculation units that can be independently operated, and each busbar segment calculation unit synchronously executes path optimization within the same clock beat; each busbar segment calculation unit, driven by a comprehensive constraint set, takes the starting connection point coordinates as input, iteratively generates a busbar segment discrete node coordinate chain, and quantitatively evaluates the busbar segment discrete node coordinate chain through an embedded busbar segment path evaluation model to obtain a busbar segment generation value, and then all busbar segment generation values are aggregated to form a topology total generation cost; the busbar segment path evaluation model takes the minimum bending radius criterion as the core and maps the internal invisible nodes of the busbar segment to the external explicit nodes; A parallel busbar level collaborative optimization framework of the target substation is built; the busbar level collaborative optimization framework is sequentially connected by a constraint injection module, the insulating tubular busbar topology mapping framework and a topology total generation cost decision module; the constraint injection module is used to issue the comprehensive constraint set and the starting connection point coordinates to the insulating tubular busbar topology mapping framework; the topology total generation cost decision module is used to collect and output the topology total generation cost of all busbar segment calculation units; The parallel insulating tubular busbar topology mapping framework comprises: According to the primary wiring diagram of the target substation, the busbar segment equivalent network corresponding to each busbar segment calculation unit is established; For each busbar segment equivalent network, the corresponding busbar segment path evaluation model is constructed; A node coordinate distributor and a generation value collector are configured for each busbar segment calculation unit; the node coordinate distributor is used to solve and generate a busbar segment discrete node coordinate chain under the driving of the comprehensive constraint set and broadcast it to all busbar segment path evaluation models in the busbar segment calculation unit as a reference, and the generation value collector is used to collect the busbar segment generation value output by all busbar segment path evaluation models in the busbar segment calculation unit and send the result to the topology total generation cost decision module.

[0006] Further, the mapping of the internal invisible node to the external explicit node in the bus section based on the minimum bending radius criterion includes: The minimum bending radius criterion is used to equivalently contract the discrete node coordinate chain in the bus section path evaluation model, eliminate the internal invisible node coordinates, and only keep the external explicit node coordinates. The bus section generation value expression of the bus section path evaluation model is: E = a·L + b·R + g·S In the formula, E represents the bus section generation value, L represents the bus section straight length, R represents the bus section bending angle, S represents the number of support insulators, a, b, and g are respectively the straight length weight coefficient, the bending angle weight coefficient, and the support insulator weight coefficient. The external explicit node generation value expression after the equivalent contraction is: Eext = a·Leq + b·Req + g·Seq In the formula, Leq represents the equivalent straight length, Req represents the equivalent bending angle, Seq represents the equivalent number of support insulators, and Eext represents the bus section generation value of the external explicit node.

[0007] Further, the construction of the corresponding bus section path evaluation model includes: The node decompression submodule, the internal state updating submodule, and the model compression submodule of the bus section path evaluation model are established; the node decompression submodule is used to read the discrete node coordinate chain of the bus section in the last iteration step, and restore all node coordinates in the equivalent network of the bus section; the internal state updating submodule is used to calculate the force history of the support insulator, the displacement history of the expansion joint, and the state history of the grounding switch in parallel; and the model compression submodule is used to aggregate the internal calculation results and output the bus section generation value.

[0008] Further, the establishment of the bus section equivalent network corresponding to each bus section calculation unit includes: Based on the Bezier curve control point mapping method, the path curve in the bus section equivalent network is established based on the bending radius constraint. And / or, based on the node coordinate invariance principle, the straight line equivalent path of the support insulator position in the bus section equivalent network is established.

[0009] Further, the method is realized based on a GIS platform or a BIM platform. When the method is realized based on the GIS platform, the multi-core CPU parallel architecture is used to realize hardware parallelism between different layers and between different cores in the same layer. When the method is implemented based on a BIM platform, an insulating tube type bus topology mapping framework and each module of a bus level collaborative optimization framework are packaged as sub-functions, and a top-level scheduling function is created as an interactive interface of the BIM program and external interfaces; the top-level scheduling function is directly connected to all external interfaces, coordinates data flow of each module, and controls execution sequences of each module.

[0010] Further, the comprehensive constraint set includes a minimum bending radius constraint, a support insulator spacing constraint, and a grounding switch position constraint. The generation step of the minimum bending radius constraint includes reading an insulating tube type bus specification database, extracting a minimum allowable bending radius corresponding to the type, and writing the minimum allowable bending radius into a global constraint register; the bus segment path evaluation model calls the global constraint register at each time of generating a new node coordinate, and if the new node coordinate causes a path segment bending radius to be less than the minimum allowable bending radius, the bus segment path evaluation model reverts to a previous node and regenerates a candidate node coordinate. The generation step of the support insulator spacing constraint includes reading a substation civil structure database and extracting a complete list of embedded part coordinates; the bus segment path evaluation model inserts a new node coordinate in the middle and forces alignment with the nearest embedded part coordinate when generating a bus segment discrete node coordinate chain, if a spacing between adjacent nodes is greater than a maximum allowable spacing. The generation step of the grounding switch position constraint includes reading a primary equipment arrangement database and extracting a list of grounding switch installation area boundaries; the bus segment path evaluation model eliminates a node coordinate and regenerates a candidate node coordinate when generating a bus segment discrete node coordinate chain, if the node coordinate is located outside any installation area boundary.

[0011] Further, the generation of the bus segment discrete node coordinate chain includes: Taking the minimum bending radius constraint, the support insulator spacing constraint, and the grounding switch position constraint as input conditions; Taking a starting connection point coordinate as a starting point, an improved A-star algorithm is called to calculate an initial bus segment discrete node coordinate chain, wherein a heuristic function is a sum of a Euclidean distance and a constraint penalty term; Three times of spline smoothing are performed on the initial bus segment discrete node coordinate chain to obtain a smoothed bus segment discrete node coordinate chain; A three-dimensional collision model of the substation is read, voxel collision detection is performed on each node coordinate, if a node coordinate intersects with an obstacle, a conflict node is marked, after the conflict node is eliminated, an improved A-star algorithm is called to regenerate a compensation node coordinate, until a collision-free bus segment discrete node coordinate chain is obtained; The collision-free bus segment discrete node coordinate chain is broadcast to the bus segment path evaluation model.

[0012] Further, the calculation of the bus segment generation value includes: Reading the insulating tube type busbar material database, extracting the unit length weight attribute, calculating the busbar segment straight line length cost; Reading the bending loss coefficient table, extracting the unit angle bending loss, calculating the busbar segment bending angle cost; Reading the support insulator cost table, extracting the installation cost of a single support insulator, counting the number of support insulators, and calculating the installation cost; The straight line length cost, the bending angle cost and the installation cost are input into the dynamic weight unit, the dynamic weight unit automatically outputs the weight coefficient according to the voltage grade of the substation, and the busbar segment value is obtained by weighted summation.

[0013] Further, the identification and removal of redundant nodes include: Traverse each node coordinate in the busbar segment equivalent network, calculate the included angle value formed by the node coordinate and the front and rear node coordinates, and convert the included angle value into a curvature value; Compare the curvature value with the preset curvature threshold value, if the curvature value is less than the preset curvature threshold value, mark the node coordinate as a redundant node; Delete the redundant node coordinate, update the connection relationship of the front and rear node coordinates, and write into the node linked list of the busbar segment equivalent network.

[0014] Further, the output of the topology total cost decision module includes: Reading the substation voltage grade and short circuit capacity, automatically generating a weight coefficient table; Inputting the busbar segment value of each busbar segment calculation unit into the weight coefficient table to obtain the corresponding weight coefficient; Multiplying the busbar segment value by the weight coefficient and accumulating to obtain the topology total cost; Inputting the topology total cost into the comparator and comparing it with the preset value threshold; If the topology total cost is less than the preset value threshold, the path output unit is called to write the current busbar segment discrete node coordinate chain into the optimized path file; If the topology total cost is not less than the preset value threshold, the constraint adjustment unit is called to automatically relax the minimum bending radius constraint, increase the support insulator spacing or expand the grounding switch installation area according to the exceeding amplitude, and return to the constraint condition generation step, the busbar segment discrete node coordinate chain generation step, the busbar segment value calculation step and the redundant node identification step until the topology total cost is less than the preset value threshold.

[0015] The above embodiments of the present application have at least the following beneficial effects: 1. By parallelly building an insulating tube type busbar topology mapping framework, the busbar path optimization problem is decomposed into multiple busbar segment calculation units that can be independently operated, realizing synchronous path optimization within the same clock beat. This parallel processing method greatly improves the calculation efficiency of path optimization, effectively solves the problem of low planning efficiency caused by relying on artificial experience in traditional methods, and can quickly generate a busbar path scheme that meets the complex substation layout requirements.

[0016] 2. A busbar segment path evaluation model with the minimum bending radius criterion as the core is adopted, the internal invisible nodes of the busbar segment are mapped to external explicit nodes, and multiple factors such as straight line length, bending angle and support insulator quantity are combined for quantitative evaluation. This model can fully reflect the actual operation state of the busbar, ensure the scientificity and rationality of path planning, effectively solve the problem of single consideration of factors in path planning in the prior art, which cannot accurately reflect the operation state of the busbar, and improve the stability and reliability of the busbar path.

[0017] 3. Through the busbar level collaborative optimization framework, dynamic adjustment of the comprehensive constraint set and real-time decision of the total topology cost are realized. During the path optimization process, when the total topology cost does not meet the preset threshold, the minimum bending radius constraint can be automatically relaxed, the support insulator spacing can be increased, or the grounding switch installation area can be expanded, and the path optimization is re-performed until the optimal solution is reached. This dynamic adjustment mechanism effectively solves the limitations of the prior art in handling complex constraint conditions, realizes the balance of path optimization under multiple constraint conditions, and improves the adaptability and flexibility of substation insulating tube type busbar path optimization. BRIEF DESCRIPTION OF DRAWINGS

[0018] The above and other objects, features and advantages of the exemplary embodiments of the present application will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: Figure 1 The flowchart of the substation insulating tube type busbar path optimization method provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0019] The technical solutions in the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, not all embodiments. The components of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0020] As shown in Figure 1 The present application provides a substation insulating tube type busbar path optimization method, which includes: S1, parallelly build an insulating tube type bus topology mapping framework of the target transformer substation; the insulating tube type bus topology mapping framework comprises a plurality of bus section calculation units capable of independent operation, each bus section calculation unit synchronously performs path optimization in the same clock beat; each bus section calculation unit, driven by a comprehensive constraint set, takes a starting connection point coordinate as input, iteratively generates a bus section discrete node coordinate chain, and quantitatively evaluates the bus section discrete node coordinate chain through an embedded bus section path evaluation model to obtain a bus section generation value, and then aggregates all bus section generation values to form a topology total generation cost; the bus section path evaluation model takes the minimum bending radius criterion as the core and maps the internal invisible node of the bus section to the external explicit node; S2, parallelly build a bus level collaborative optimization framework of the target transformer substation; the bus level collaborative optimization framework is sequentially connected by a constraint injection module, the insulating tube type bus topology mapping framework and a topology total generation cost decision module; the constraint injection module is used to issue the comprehensive constraint set and the starting connection point coordinate to the insulating tube type bus topology mapping framework; the topology total generation cost decision module is used to collect and output the topology total generation cost of all bus section calculation units.

[0021] The parallelly built insulating tube type bus topology mapping framework comprises: According to the primary wiring diagram of the target transformer substation, establish the bus section equivalent network corresponding to each bus section calculation unit; For each bus section equivalent network, build a corresponding bus section path evaluation model; Configure a node coordinate distributor and a generation value collector for each bus section calculation unit; the node coordinate distributor is used to solve and generate a bus section discrete node coordinate chain under the driving of the comprehensive constraint set and broadcast it to all bus section path evaluation models in the bus section calculation unit, and the generation value collector is used to collect the bus section generation value output by all bus section path evaluation models in the bus section calculation unit and send the result to the topology total generation cost decision module.

[0022] The insulation pipe type busbar topology mapping framework refers to converting the busbar structure of the target substation into a topological model that can be calculated in parallel. Specifically, a multi-threaded or distributed computing architecture can be used to achieve this. By decomposing the busbar into multiple independent busbar segment calculation units, parallel processing of path optimization is achieved, thereby improving computational efficiency. The busbar segment calculation unit refers to an operation module that can independently perform path optimization. It can be implemented using threads or processes in a computer program. Each unit performs calculations synchronously within the same clock cycle, ensuring that the path generation process for each busbar segment maintains temporal consistency and avoids path conflicts caused by timing differences. The comprehensive constraint set refers to a set of physical and electrical conditions that must be followed in the design of the busbar path. It can be stored in the form of a database or configuration file. By injecting constraint conditions uniformly into each busbar segment calculation unit, it ensures that the path generation process always complies with safety specifications.

[0023] The busbar segment discrete node coordinate chain refers to a sequence of busbar path nodes composed of multiple discrete point coordinates. It can be implemented using coordinate chain tables or array data structures. By iteratively generating node coordinates and connecting them into a chain, a continuous busbar path trajectory is formed. The busbar segment path evaluation model is a mathematical model used to quantitatively evaluate the quality of the busbar path. It can be implemented using a weighted scoring algorithm. By introducing the minimum bending radius criterion, the geometric characteristics of the path are converted into a value, providing a quantitative basis for path optimization. The busbar-level collaborative optimization framework is a system architecture that coordinates the global optimization of each busbar segment calculation unit. It can be implemented using a modular software design method. By using a constraint injection module to uniformly distribute constraint conditions, and a topology total cost decision module to aggregate the value of each unit, the global path cost is evaluated and decided.

[0024] The substation insulation pipe type busbar path trajectory optimization method includes two main steps: parallel construction of the insulation pipe type busbar topology mapping framework and the busbar-level collaborative optimization framework.

[0025] The insulation pipe type busbar topology mapping framework includes multiple busbar segment calculation units that can be independently calculated. Each busbar segment calculation unit performs path optimization synchronously within the same clock cycle. Each busbar segment calculation unit, driven by the comprehensive constraint set, takes the starting connection point coordinates as input and iteratively generates a busbar segment discrete node coordinate chain. The busbar segment path evaluation model is embedded in the busbar segment discrete node coordinate chain to quantitatively evaluate the busbar segment value. Subsequently, all busbar segment values are aggregated to form a topology total cost. The busbar segment path evaluation model uses the minimum bending radius criterion as the core, mapping the internal invisible nodes of the busbar segment to external explicit nodes.

[0026] The bus level collaborative optimization framework is sequentially connected by a constraint injection module, an insulated tube type bus topology mapping framework, and a topology total cost decision module. The constraint injection module issues a comprehensive constraint set and a starting connection point coordinate to the insulated tube type bus topology mapping framework. The topology total cost decision module collects and outputs the topology total cost of all bus segment calculation units.

[0027] The parallel construction of the insulated tube type bus topology mapping framework includes the following steps: first, according to the target substation primary wiring diagram, the bus segment equivalent network corresponding to each bus segment calculation unit is established. Then, for each bus segment equivalent network, the corresponding bus segment path evaluation model is constructed. The node coordinate distributor and the generation value collector are configured for each bus segment calculation unit. The node coordinate distributor solves and generates the bus segment discrete node coordinate chain based on the starting connection point coordinate as the reference under the driving of the comprehensive constraint set, and broadcasts it to all bus segment path evaluation models in the bus segment calculation unit. The generation value collector collects the bus segment generation value output by all bus segment path evaluation models in the bus segment calculation unit and sends the result to the topology total cost decision module.

[0028] After the parallel construction of the insulated tube type bus topology mapping framework, it also includes identifying and removing redundant nodes in each bus segment equivalent network.

[0029] The method improves the calculation efficiency through the parallel architecture, realizes the real-time feedback of the path evaluation through the dynamic constraint injection, and ensures the smoothness of the path curve through the removal of redundant nodes. The introduction of the minimum bending radius criterion effectively avoids the stress concentration problem. The collaborative optimization of multiple bus segments improves the optimality of global path planning.

[0030] As a preferred embodiment, the scheme of the application is implemented as follows: Taking a 500kV intelligent substation as an example, first, an insulated tube type bus topology mapping framework is established. The framework includes 6 independent bus segment calculation units corresponding to 6 main bus sections in the substation. Each calculation unit is equipped with a node coordinate distributor and a generation value collector.

[0031] The node coordinate distributor receives the comprehensive constraint set, including the minimum bending radius of 50cm, the support insulator spacing of 5m, the installation area boundary of the grounding switch, and other constraint conditions. The initial discrete node coordinate chain is generated based on the starting connection point coordinate by using the improved A* algorithm.

[0032] The bus segment path evaluation model uses the minimum bending radius criterion to map the internal invisible nodes to the external explicit nodes. The evaluation model calculates the straight line length cost, the bending angle cost, and the support insulator installation cost, and obtains the bus segment generation value by weighted summation according to the dynamic weight coefficient.

[0033] The generation value collector aggregates the generation values of the 6 bus section calculation units and sends them to the topology total generation cost decision module. The decision module compares the topology total generation cost with a preset threshold value. If the topology total generation cost is less than the threshold value, the optimized path is output. Otherwise, the constraint condition is adjusted and the iteration is restarted.

[0034] Finally, the redundant nodes are identified and removed. The curvature value corresponding to the included angle of adjacent nodes is calculated. If the curvature value is less than a preset threshold value 0.01, the redundant node is marked and deleted, and the node connection relationship is updated.

[0035] The entire optimization process is executed in parallel on a multi-core CPU, and the single iteration time is controlled within 1 minute. Through multiple iterations, the bus path scheme that satisfies all constraint conditions and has the minimum total cost is finally obtained.

[0036] The application further proposes a bus section path evaluation model with the minimum bending radius criterion as the core, which maps the internal invisible nodes of the bus section to external explicit nodes, including: using the minimum bending radius criterion to equivalently contract the discrete node coordinate chain of the bus section, eliminating the internal invisible node coordinates, and only retaining the external explicit node coordinates; the generation value expression of the bus section is E = a·L + β·R + γ·S, and the generation value expression of the external explicit node after equivalent contraction is Eext = a·Leq + β·Req + γ·Seq.

[0037] The equivalent contraction operation simplifies the continuous curved path into an equivalent combination of straight line segments and polyline segments through geometric transformation. The minimum bending radius criterion is converted into a curvature constraint condition, which acts on the curvature radius calculation process of the node coordinate chain. The parameters in the generation value expression are dynamically adjusted through weight coefficients, for example, the a coefficient is set according to the tensile strength of the bus material, the β coefficient is calibrated according to the bending loss experimental data, and the γ coefficient is associated with the purchase cost of the support insulator. The Leq value after equivalent contraction is calculated by the original path length through the piecewise integral method, the Req value is generated by the weighted average of the reciprocal of the curvature radius, and the Seq value is automatically converted according to the support point density.

[0038] Specifically, in the path optimization process, after the bus segment discrete node coordinate chain is generated, firstly, the curvature of the path segment formed by adjacent nodes is detected. If the curvature radius of a certain path segment is greater than the minimum allowable bending radius, the path segment is equivalent to a straight line segment, Leq takes the actual length, and Req is set to zero; if the curvature radius is equal to the minimum allowable bending radius, an equivalent polyline segment is generated, Leq takes the chord length, and Req takes the actual bending angle. The number of support insulators Seq is obtained by rounding up the ratio of the equivalent path segment length to the maximum allowable spacing. This equivalent mapping process makes the value calculation only need to process the explicit nodes, and the calculation amount is reduced to 23.7% of the original method. For example, when the original path contains 15 invisible nodes, only 3 explicit nodes are retained after equivalent contraction, and the value calculation time is shortened to 18.5% of the original process. The weight coefficients α, β and γ are optimized by orthogonal test method, and the typical values are α=0.6, β=0.3 and γ=0.1, which ensure that the influence of bending loss on the value is controlled within the range of 30%±5%.

[0039] As a preferred embodiment, the scheme of the application is implemented as follows: The minimum bending radius criterion is used to equivalently contract the bus segment discrete node coordinate chain in the bus segment path evaluation model, to eliminate the internal invisible node coordinates and only retain the external explicit node coordinates. Specifically, first, the bus segment discrete node coordinate chain is obtained, which includes multiple node coordinates. Then, according to a preset minimum bending radius value, the curvature between adjacent nodes is calculated. If the calculated curvature is less than the curvature corresponding to the minimum bending radius, the node coordinate is retained; otherwise, the node is marked as an internal invisible node. Next, the coordinates marked as internal invisible nodes are deleted, and only the external explicit node coordinates are retained. Finally, the retained external explicit node coordinates are reconnected to form the equivalently contracted bus segment path.

[0040] Further, the bus segment value expression of the bus segment path evaluation model is: E=α·L+β·R+γ·S In the formula, E represents the bus segment value, L represents the bus segment straight line length, R represents the bus segment bending angle, S represents the number of support insulators, and α, β and γ are respectively the straight line length weight coefficient, the bending angle weight coefficient and the support insulator weight coefficient.

[0041] Therefore, the external explicit node value expression of the equivalently contracted bus segment is: Eext=α·Leq+β·Req+γ·Seq In the formula, Leq represents the equivalent straight line length, Req represents the equivalent bending angle, Seq represents the equivalent number of support insulators, and Eext represents the bus segment value of the external explicit node.

[0042] In a specific implementation, first, the equivalent straight line length Leq is calculated, that is, the straight line distance is accumulated for the coordinates of the remaining external explicit nodes. Second, the equivalent bending angle Req is calculated, that is, the included angle between adjacent external explicit nodes is accumulated. Then, the equivalent support insulator quantity Seq is counted, that is, the nodes in the remaining external explicit nodes that need to be installed with support insulators are counted. Finally, the calculated Leq, Req, and Seq are substituted into the equivalent contracted external explicit node value expression to obtain the Eext value.

[0043] The application further proposes a node decompression submodule, an internal state updating submodule, and a model compression submodule for establishing a bus section path evaluation model. The node decompression submodule reads the bus section discrete node coordinate chain of the previous iteration step to restore all node coordinates inside the equivalent network of the bus section; the internal state updating submodule calculates the support insulator force history, the expansion joint displacement history, and the grounding switch state history in parallel; and the model compression submodule aggregates the internal calculation results and outputs the bus section value.

[0044] The node decompression submodule restores the discrete node sequence to a complete set of spatial coordinates through a coordinate chain analysis algorithm, ensuring the continuity of the initial state of each iteration with the previous period. The internal state updating submodule uses a distributed computing unit to perform time-domain integration of the axial force and bending moment of the support insulator, difference calculation of the expansion joint displacement, and statistical analysis of the contact pressure of the grounding switch contact. The model compression submodule converts multi-dimensional physical quantities into dimensionless values through data normalization processing and outputs a scalar result using a weighted summation method.

[0045] Specifically, the node decompression submodule reads the coordinate chain data in binary format from the storage unit at the beginning of each iteration and restores the three-dimensional coordinate point set through a floating-point number decoder. In the internal state updating submodule, the support insulator force calculation unit calls a finite element solver to calculate the stress distribution based on the node coordinate change; the expansion joint displacement calculation unit uses the Lagrange interpolation method to track the relative motion of adjacent nodes; and the grounding switch state calculation unit evaluates the connection reliability through a contact resistance model. The model compression submodule maps each physical quantity to a unified dimension according to a preset conversion coefficient and generates the final value after linear combination. This structure effectively avoids potential mechanical failure risks during path optimization by updating mechanical state parameters in real time while maintaining computational efficiency.

[0046] As a preferred embodiment, the scheme of the application is implemented as follows: When constructing the bus segment path evaluation model, the node decompression submodule, the internal state update submodule, and the model compression submodule are established. The node decompression submodule reads the discrete node coordinate chain of the bus segment in the last iteration step to restore all node coordinates in the equivalent network of the bus segment. The internal state update submodule calculates the force history of the support insulator, the displacement history of the expansion joint, and the state history of the grounding switch in parallel. The model compression submodule aggregates the internal calculation results and outputs the bus segment value.

[0047] Specifically, the node decompression submodule uses a hash table to store node coordinate information, with node number as the key and coordinate value as the value. When receiving a new discrete node coordinate chain of the bus segment, each node in the coordinate chain is traversed, and the corresponding coordinate value in the hash table is updated. For nodes not in the coordinate chain, their original coordinate values are retained.

[0048] The internal state update submodule uses a multi-thread parallel calculation method. The force history of the support insulator is calculated by the finite element analysis method, considering factors such as the self-weight of the bus and wind load. The displacement history of the expansion joint is calculated based on the thermal expansion coefficient and temperature change. The state history of the grounding switch is simulated according to the preset operation period.

[0049] The model compression submodule uses the weighted summation method to aggregate the calculation results. The support insulator force, expansion joint displacement, and grounding switch state are assigned different weights, and the bus segment value is obtained by comprehensive calculation. The value is standardized to ensure the comparability between different bus segments.

[0050] The application further proposes a Bezier curve control point mapping method to establish a path curve in the equivalent network of the bus segment, with bending radius constraint as the core; and / or, based on the node coordinate invariable principle, a straight line segment equivalent path of the support insulator position in the equivalent network of the bus segment.

[0051] The Bezier curve control point mapping method defines a cubic Bezier curve by selecting four control points. The first and last control points are fixed at the path endpoints, and the middle two control points are symmetrically distributed along the bus direction with a spacing of twice the minimum bending radius. The node coordinate invariable principle locks the three-dimensional coordinates of the node where the support insulator is located, maintains the absolute position of adjacent straight line segments in the equivalent network unchanged, and only allows non-support nodes to adjust within the constraint range.

[0052] Specifically, when processing the curved section, a cubic Bezier curve control point mapping method is adopted, and by adjusting the position of the intermediate control point, it is ensured that the curvature radius of the generated path curve is always greater than or equal to the minimum bending radius specified by the material. For example, for the Φ200mm insulating tube type busbar commonly used in 220kV substations, the intermediate control point spacing is set to 1.2 meters, corresponding to a minimum bending radius of 0.6 meters. When processing the straight section where the supporting insulator is located, after fixing the support point coordinates, an equivalent path is generated by linear interpolation to ensure that the spacing error of the supporting insulator is controlled within ±5mm. The two methods can be applied independently or in combination, and when used simultaneously, the control points of the curved section and the support points of the straight section are seamlessly connected through a coordinate transformation matrix to form a complete equivalent network topology.

[0053] As a preferred embodiment, the scheme of the present application is implemented as follows: When establishing the path curve in the busbar section equivalent network with bending radius constraint as the core, the Bezier curve control point mapping method is adopted. Specifically, first, the starting point and ending point coordinates of the busbar section are determined, and then a plurality of control points are selected between the two points. The positions of these control points are determined by the bending radius constraint, for example, for an insulating tube type busbar with a minimum bending radius of 2 meters, the selection of control points needs to ensure that the curvature radius of the curve at any position is not less than 2 meters. Next, the discrete point coordinates on the path are calculated using the Bezier curve formula to form a smooth and continuous curve path.

[0054] Further, when establishing the straight section equivalent path of the supporting insulator position in the busbar section equivalent network, the node coordinate invariance principle is used. In specific implementation, first, the installation position coordinates of the supporting insulator are determined, and these coordinates are used as fixed nodes of the path. Then, straight sections are generated between adjacent fixed nodes, and these straight sections constitute the equivalent path of the supporting insulator position. In this way, the accuracy and stability of the supporting insulator position are ensured.

[0055] The present application further proposes that the method is implemented based on a GIS platform or a BIM platform; when implemented based on the GIS platform, a multi-core CPU parallel architecture is adopted to realize hardware parallelism between different layers and between different cores of the same layer; when implemented based on the BIM platform, the modules of the insulating tube type busbar topology mapping framework and the busbar level collaborative optimization framework are packaged as sub-functions, and a top-level scheduling function is created as the interaction interface of the BIM program and external interfaces; the top-level scheduling function directly connects all external interfaces, coordinates the data flow of each module, and controls the execution order of each module.

[0056] In the GIS platform-based implementation, the multi-core CPU parallel architecture divides the three-dimensional spatial data of the substation into multiple independent layers, each of which is assigned to a different computing core for path optimization operation. Multiple bus segment calculation units within the same layer achieve data synchronization through shared memory, and cross-core communication between different layers is achieved through a message passing interface. In the BIM platform-based implementation, the sub-function encapsulation process converts the constraint injection, node coordinate distribution, and generation value collection operations involved in bus path optimization into independent function units. The top-level scheduling function triggers the execution of each sub-function through an event-driven mechanism and establishes a real-time data channel with the BIM model database.

[0057] Specifically, when using the GIS platform, the substation geographic information data is divided into an electrical equipment layer, a civil structure layer, and a collision detection layer, each of which is assigned to an independent CPU core for operation. The bus segment calculation unit performs path generation in the electrical equipment layer, and real-time calls the embedded coordinate data of the civil structure layer and performs conflict verification through the voxel model of the collision detection layer. The multi-core parallel architecture enables synchronous advancement of data calling and computing tasks. When using the BIM platform, the encapsulated sub-functions receive external input constraints and device coordinates through a parameterized interface. The top-level scheduling function sequentially calls the constraint injection module to generate a comprehensive constraint set according to the path optimization process, triggers the iteration calculation of the insulated tubular bus topology mapping framework, and finally returns the resolved topology total cost to the BIM model visualization interface. This layered encapsulation architecture enables the BIM platform to adapt to different substation engineering scenarios by adjusting the execution logic of the top-level scheduling function without modifying the core algorithm.

[0058] As a preferred embodiment, the scheme of the present application is implemented as follows: The substation insulated tubular bus path optimization method is implemented based on a GIS platform or a BIM platform. In the GIS platform-based implementation, a multi-core CPU parallel architecture is used to achieve hardware parallelism between different layers and different cores within the same layer. In the BIM platform-based implementation, the insulated tubular bus topology mapping framework and the modules of the bus-level collaborative optimization framework are encapsulated as sub-functions, and a top-level scheduling function is created as an interactive interface between the BIM program and external interfaces. The top-level scheduling function directly connects all external interfaces, coordinates the data flow of each module, and controls the execution order of each module.

[0059] Specifically, when implemented based on the GIS platform, an Intel Xeon E5-2699 v4 processor can be used, which has 22 physical cores and can support 44 logical cores through the hyper-threading technology. The busbar systems of different voltage levels of the substation are distributed to different GIS layers, and each layer is processed by an independent CPU core. For example, the 500 kV, 220 kV and 110 kV busbar systems are processed by 3 independent physical cores respectively. At the same time, within each voltage level layer, the busbar segment calculation units are distributed to different logical cores to realize parallel calculation within the same layer.

[0060] When implemented based on the BIM platform, Autodesk Revit can be used as the basic platform, and plug-ins are developed through Revit API. Each functional module of the insulated pipe-type busbar topology mapping framework and the busbar level collaborative optimization framework is encapsulated as an independent C# class, each class containing corresponding methods and attributes. For example, the TopologyMapper class is created to encapsulate the topology mapping function, and the Optimization Coordinator class is created to encapsulate the collaborative optimization function. Then a Main Scheduler class is created as the top-level scheduling function, which contains methods such as Initialize Parameters(), Run Optimization() and OutputResults(), respectively used for initializing parameters, executing the optimization process and outputting results. The Main Scheduler class interacts with the Revit program through the External Command interface of Revit API, receives the substation parameters and constraint conditions input by the user, and feeds back the optimization results directly to the three-dimensional model of Revit.

[0061] The application further proposes that the comprehensive constraint set comprises a minimum bending radius constraint, a support insulator spacing constraint and a grounding switch position constraint; the generation step of the minimum bending radius constraint comprises reading an insulating tube type busbar specification database, extracting a minimum allowable bending radius corresponding to the type, and writing into a global constraint register; the busbar segment path evaluation model calls the global constraint register at each time of generating a new node coordinate, and if the new node coordinate causes the path segment bending radius to be less than the minimum allowable bending radius, the last node is rolled back and the candidate node coordinate is regenerated; the generation step of the support insulator spacing constraint comprises reading a power substation civil structure database and extracting a complete embedded part coordinate list; the busbar segment path evaluation model, when generating a busbar segment discrete node coordinate chain, if the spacing between adjacent nodes is greater than the maximum allowable spacing, a new node coordinate is inserted in the middle and forced to align with the nearest embedded part coordinate; the generation step of the grounding switch position constraint comprises reading a primary equipment arrangement database and extracting a grounding switch installation area boundary list; the busbar segment path evaluation model, when generating a busbar segment discrete node coordinate chain, if the node coordinate is located outside any installation area boundary, the node coordinate is removed and the candidate node coordinate is regenerated.

[0062] The minimum bending radius constraint ensures that the bending radius always meets the material mechanics requirements in the path generation process by dynamically calling the parameters in the insulating tube type busbar specification database; the support insulator spacing constraint ensures the reasonable distribution of the support structure by forced alignment of the embedded part coordinates; and the grounding switch position constraint avoids equipment installation conflicts through boundary detection. The generation processes of the three constraints are all based on real-time data reading and updating of the database, forming a closed-loop feedback mechanism.

[0063] Specifically, in the path optimization process, the busbar segment path evaluation model first obtains the minimum allowable bending radius from the global constraint register, and detects the curvature of the path segment in real time when generating a new node coordinate. If it is detected that the bending radius does not meet the requirements, the node rollback and regeneration mechanism is triggered to avoid the occurrence of material stress over-limit problem in path planning. The support insulator spacing constraint inserts a new node when the spacing between adjacent nodes exceeds the threshold value through matching of the embedded part coordinate list, to ensure the accurate alignment of the support point and the civil embedded structure and reduce the later construction adjustment cost. The grounding switch position constraint automatically removes the node coordinates that exceed the allowed range through spatial screening of the installation area boundary, to eliminate the risk of equipment arrangement conflicts. Through the synergistic effect of the above constraints, the path planning process can realize the global optimization of the busbar layout under the premise of meeting the multi-dimensional engineering requirements, and improve the feasibility and economy of the planning results.

[0064] As a preferred embodiment, the scheme of the application is implemented as follows: The comprehensive constraint set comprises a minimum bending radius constraint, a support insulator spacing constraint and a grounding switch position constraint.

[0065] The generation step of the minimum bending radius constraint includes reading the insulated tube busbar specification database, extracting the minimum allowed bending radius for the corresponding model, and writing it to the global constraint register. The busbar segment path evaluation model calls the global constraint register every time a new node coordinate is generated. If the new node coordinate results in a path segment bending radius less than the minimum allowed bending radius, the system reverts to the previous node and regenerates the candidate node coordinate.

[0066] The generation step of the support insulator spacing constraint includes reading the substation civil structure database and extracting the list of all embedded component coordinates. The busbar segment path evaluation model generates a discrete node coordinate chain for the busbar segment. If the spacing between adjacent nodes is greater than the maximum allowed spacing, the system inserts a new node coordinate in the middle and forces it to align with the nearest embedded component coordinate.

[0067] The generation step of the grounding switch position constraint includes reading the primary equipment arrangement database and extracting the list of grounding switch installation area boundaries. The busbar segment path evaluation model generates a discrete node coordinate chain for the busbar segment. If a node coordinate is outside any of the installation area boundaries, the system removes that node coordinate and regenerates a candidate node coordinate that satisfies the condition.

[0068] For example, in the process of optimizing the path of an insulated tube busbar in a substation, the minimum allowed bending radius is first read from the insulated tube busbar specification database, which is 2 meters. Then, the substation civil structure database is read to obtain the list of embedded component coordinates, which includes 100 embedded component coordinates. Next, the primary equipment arrangement database is read to obtain the boundary coordinates of 5 grounding switch installation areas.

[0069] When generating a discrete node coordinate chain for the busbar segment, the system first checks whether each newly generated node coordinate satisfies the minimum bending radius constraint. If it does not, the system reverts to the previous node and regenerates a candidate node coordinate until the constraint is satisfied.

[0070] Further, the system checks the distance between adjacent nodes. Assuming the maximum allowed spacing is 5 meters, if the distance between adjacent nodes exceeds 5 meters, the system inserts a new node coordinate in the middle and aligns it with the nearest embedded component coordinate.

[0071] Finally, the system checks whether each node coordinate is within the grounding switch installation area. If a node coordinate is found to be outside the installation area boundary, the system removes that node coordinate and regenerates a candidate node coordinate that satisfies the condition.

[0072] The application further provides a generation method of bus segment discrete node coordinate chain, comprising taking minimum bending radius constraint, support insulator spacing constraint and grounding switch position constraint as input conditions; taking starting connection point coordinates as a starting point, calling an improved A-star algorithm to calculate initial bus segment discrete node coordinate chain, wherein a heuristic function is a sum of Euclidean distance and constraint penalty term; performing three times of spline smoothing on the initial bus segment discrete node coordinate chain to obtain smoothed bus segment discrete node coordinate chain; reading a three-dimensional collision model of a substation, performing voxel collision detection on node coordinates, marking a conflict node if the node coordinates intersect with an obstacle, calling the improved A-star algorithm to regenerate compensation node coordinates after removing the conflict node, and obtaining a non-collision bus segment discrete node coordinate chain until a non-collision bus segment discrete node coordinate chain is obtained; and broadcasting the non-collision bus segment discrete node coordinate chain to a bus segment path evaluation model.

[0073] The heuristic function of the improved A-star algorithm is composed of linear superposition of Euclidean distance and constraint penalty term, and the constraint penalty term dynamically adjusts the weight according to the degree of violation of the minimum bending radius, support insulator spacing or grounding switch position constraint by the node coordinates. The three times of spline smoothing adopts a piecewise cubic polynomial interpolation method to eliminate the jagged node distribution in the initial path under the premise of maintaining the continuity of the path curvature. The voxel collision detection is based on the spatial segmentation data of the three-dimensional model of the substation, and performs intersection operation on the obstacle voxel after converting the bus path node coordinates into voxel coordinates, and the collision detection accuracy is controlled by the voxel resolution.

[0074] Specifically, the initial path generation stage finds a feasible path under the conditions of satisfying the bending radius, insulator spacing and other hard constraints by using the improved A-star algorithm, wherein the constraint penalty term quantifies the degree of deviation of the path from the constraint condition as a generation value, and the path length together constitutes the path evaluation standard. The three times of spline smoothing stage performs interpolation processing on the nodes of the initial path, and adjusts the control points so that the path curvature change rate does not exceed the mechanical stress threshold allowed by the bus material. The collision detection stage adopts a layered detection strategy, first performs coarse voxel detection to quickly exclude obvious conflict nodes, and then performs fine grid detection on the suspected conflict area. In the compensation node generation process, the improved A-star algorithm searches again in the remaining path space after removing the conflict nodes, and balances the path optimization efficiency and accuracy by dynamically adjusting the search step length. The finally generated bus segment discrete node coordinate chain not only satisfies the geometric constraint conditions, but also has engineering implementation feasibility, wherein the support insulator installation position is accurately aligned with the pre-embedded part coordinates, and the path direction in the grounding switch area meets the equipment installation specification.

[0075] As a preferred embodiment, the scheme of the application is implemented as follows: The generation of the bus segment discrete node coordinate chain comprises the following steps: Firstly, the minimum bending radius constraint, the support insulator spacing constraint, and the grounding switch position constraint are taken as input conditions. These constraints are extracted from the substation design specification and the equipment parameter database.

[0076] Secondly, the improved A-star algorithm is called to calculate the initial bus segment discrete node coordinate chain with the starting connection point coordinates as the starting point. The heuristic function of the improved A-star algorithm is set as the sum of the Euclidean distance and the constraint penalty term. The constraint penalty term is calculated according to whether the node violates the input constraint condition.

[0077] Then, the initial bus segment discrete node coordinate chain is executed three times of spline smoothing to obtain the smoothed bus segment discrete node coordinate chain. The three times of spline smoothing adopts the piecewise cubic polynomial function to ensure the continuity and smoothness between nodes.

[0078] Next, the three-dimensional collision model of the substation is read, and the voxel collision detection is performed on the node coordinates. The three-dimensional collision model of the substation contains the geometric information of all equipment, building structures, and other obstacles. The voxel collision detection discretizes the space into a cubic grid and checks whether each node coordinate intersects with the obstacle.

[0079] If the node coordinate intersects with the obstacle, it is marked as a conflict node. After removing the conflict node, the improved A-star algorithm is called again to regenerate the compensation node coordinates. This process is iterated until the collision-free bus segment discrete node coordinate chain is obtained.

[0080] Finally, the collision-free bus segment discrete node coordinate chain is broadcast to the bus segment path evaluation model for subsequent path evaluation and optimization.

[0081] The application further proposes the calculation of the bus segment value, including: reading the insulating tube type bus material database, extracting the unit length weight attribute, calculating the bus segment straight line length cost; reading the bending loss coefficient table, extracting the unit angle bending loss, calculating the bus segment bending angle cost; reading the support insulator cost table, extracting the installation cost of a single support insulator, counting the number of support insulators, and calculating the installation cost; inputting the straight line length cost, bending angle cost, and installation cost into the dynamic weight unit, and the dynamic weight unit automatically outputs the weight coefficient according to the substation voltage grade. The bus segment value is obtained by weighted summation.

[0082] The calculation of the busbar segment straight line length cost is generated by accessing the unit length weight attribute of the insulating tube type busbar material database in combination with the geometric length of the discrete node coordinate chain; the calculation of the busbar segment bending angle cost is generated by cumulatively adding the unit angle loss parameter obtained by querying the bending loss coefficient table in combination with the path bending angle; the calculation of the installation cost is generated by multiplying the unit price in the support insulator cost table based on the number of support insulators. The dynamic weight unit automatically adjusts the weight coefficient according to the voltage level of the substation, for example, the bending angle weight coefficient is increased in the high voltage level scenario to preferentially control the bending loss, and the straight line length weight coefficient is increased in the low voltage level scenario to preferentially reduce the material cost.

[0083] Specifically, the calculation process of the busbar segment value first obtains the parameters related to the physical properties of the busbar through the material database, the bending loss coefficient table and the support insulator cost table respectively, to ensure that the calculation basis is consistent with the actual engineering data. Subsequently, the dynamic weight unit dynamically adjusts the weight coefficients of each cost item according to the voltage level, for example, in a 220kV substation, the dynamic weight unit sets the bending angle weight coefficient to 0.5, the straight line length weight coefficient to 0.3, and the support insulator weight coefficient to 0.2; while in a 110kV substation, the straight line length weight coefficient is adjusted to 0.4, the bending angle weight coefficient is adjusted to 0.4, and the support insulator weight coefficient is adjusted to 0.2. Through weighted summation, the final generated busbar segment value can reflect the comprehensive influence of material cost, bending loss and installation cost under different voltage levels, thereby improving the economy and applicability of the path optimization result.

[0084] As a preferred embodiment, the scheme of the application is implemented as follows: in the calculation of the busbar segment cost value, first, the material specification parameters corresponding to the current busbar segment are extracted from the insulated tube type busbar material database, wherein the weight per unit length attribute is automatically loaded into the calculation engine. Then, according to the three-dimensional spatial position relationship of each node in the busbar segment discrete node coordinate chain, the straight line distance between adjacent nodes is accumulated by integral algorithm to generate the busbar segment straight line length cost. The bending loss coefficient table is called in real time, which stores the mapping relationship of unit angle bending loss under different voltage levels. By analyzing the actual angle values of all bending sections in the busbar segment path and superimposed calculation, the busbar segment bending angle cost is obtained. The support insulator cost table is connected with the material management system through the interface, and the installation cost of a single support insulator is dynamically obtained. At the same time, based on the number of nodes aligned with the embedded parts in the discrete node coordinate chain, the total number of support insulators is automatically counted and the installation cost is generated. After receiving the input signal of the substation voltage level, the dynamic weight unit matches the corresponding weight coefficient combination from the preset weight configuration file. When the voltage level is 220kV, the straight line length weight coefficient, the bending angle weight coefficient and the support insulator weight coefficient are set to 0.5, 0.3 and 0.2 respectively. Finally, each cost component is subjected to vector point multiplication operation with the corresponding weight coefficient, and the weighted sum is output to the cost value collector.

[0085] The application further proposes that the calculation of the busbar segment cost value comprises: reading the insulated tube type busbar material database, extracting the weight per unit length attribute, calculating the busbar segment straight line length cost; reading the bending loss coefficient table, extracting the unit angle bending loss, calculating the busbar segment bending angle cost; reading the support insulator cost table, extracting the installation cost of a single support insulator, counting the number of support insulators, calculating the installation cost; inputting the straight line length cost, the bending angle cost and the installation cost into the dynamic weight unit, and the dynamic weight unit automatically outputs the weight coefficient according to the voltage level of the substation. The weighted sum is obtained.

[0086] The mapping relationship table between the voltage level and the weight coefficient is built in the dynamic weight unit, which is generated by fitting the engineering test data. When the voltage level is 220kV, the bending angle weight coefficient is automatically increased to 0.45, and the straight line length weight coefficient is adjusted to 0.35; when the voltage level is 110kV, the straight line length weight coefficient is increased to 0.5, and the bending angle weight coefficient is reduced to 0.3. The weight coefficient of the support insulator is positively correlated with the voltage level, and the coefficient increases by 0.05 for each voltage level threshold. The material database stores the density parameters of different busbar models, and the density value of the current busbar model is called when calculating the straight line length cost.

[0087] Specifically, the material database is connected to the busbar procurement list in real time, and when the straight length cost is calculated, the system automatically matches the busbar model selected in the current project to extract the unit length weight data. The bending loss coefficient table stores the loss coefficient curves corresponding to different bending angles. After collecting the actual bending data through the angle sensor, a dynamic updated loss coefficient is generated by using a cubic polynomial fitting. The support insulator cost table integrates the supplier's quotation data, and when the number of insulators is counted, the system synchronously calculates the matching degree of the installation position and the embedded part, and automatically increases the number of insulators when the distance exceeds the threshold. After receiving the voltage level parameter of the substation, the dynamic weight unit calls the preset weight distribution algorithm, which automatically distributes the weight proportion of each cost item based on the relevance of voltage level and busbar mechanical strength and electrical insulation requirements. During the weighting process, the system multiplies the straight length cost by the alpha coefficient, the bending angle cost by the beta coefficient, and the installation cost by the gamma coefficient. The sum of the three generates a busbar segment value with voltage adaptability. This value serves as the evaluation benchmark for path optimization iteration, ensuring that high-voltage scenarios prioritize bending deformation and low-voltage scenarios focus on shortening path length, achieving optimal path selection under different working conditions.

[0088] As a preferred embodiment, the scheme of the application is implemented as follows: when performing redundant node identification in the busbar segment equivalent network, first, the current busbar segment discrete node coordinate chain is obtained, and each node coordinate in the linked list is traversed according to the storage order. For each intermediate node, the included angle value between the two adjacent line segments formed by the node and its predecessor node and successor node is calculated, wherein the predecessor node and the current node form the first line segment, and the current node and the successor node form the second line segment. The direction vectors of the two line segments are calculated through vector operation, and the cosine value of the included angle is derived using the vector dot product formula, and then converted to a curvature value. Further, the calculated curvature value is compared with a preset curvature threshold, which is set according to the bending stiffness parameter of the busbar model. When the curvature value of a node is lower than the threshold, it is determined that the node does not have a substantial impact on the path direction, and is marked as a redundant node. Then the redundant node coordinate is deleted, the predecessor node is directly connected to the successor node to form a new straight line segment, and the connection relationship of the node linked list is updated. Finally, the updated node linked list is written back to the storage unit, and the redundant node removal operation is completed.

[0089] The application further proposes that the output of the topological total cost decision module includes: reading the substation voltage level and short-circuit capacity, automatically generating a weight coefficient table; inputting the bus segment cost value of each bus segment calculation unit into the weight coefficient table to obtain the corresponding weight coefficient; multiplying the bus segment cost value and the weight coefficient and then accumulating to obtain the topological total cost; inputting the topological total cost into a comparator and comparing it with a preset cost value threshold; if the topological total cost is less than the preset cost value threshold, calling a path output unit to write the current bus segment discrete node coordinate chain into an optimized path file; if the topological total cost is not less than the preset cost value threshold, calling a constraint adjustment unit to automatically relax the minimum bending radius constraint, increase the support insulator spacing or expand the installation area of the grounding switch according to the exceeding amplitude, and returning to the constraint condition generation step, the bus segment discrete node coordinate chain generation step, the bus segment cost value calculation step and the redundant node identification step until the topological total cost is less than the preset cost value threshold.

[0090] The generation of the weight coefficient table is achieved by analyzing the substation voltage level and short-circuit capacity data to establish the mapping relationship between the voltage level and the straight line length weight coefficient, the mapping relationship between the short-circuit capacity and the bending angle weight coefficient, and the support insulator quantity weight coefficient is dynamically adjusted according to the product of the voltage level and the short-circuit capacity. The comparator adopts a segmented threshold comparison strategy, and the preset cost value threshold is divided into multiple intervals, and different intervals correspond to different constraint adjustment amplitudes. The constraint adjustment unit has a priority logic built-in, when the topological total cost exceeds the threshold, the constraint condition with the largest contribution to the cost value is relaxed first, for example, when the bending angle weight coefficient is high, the minimum bending radius constraint is adjusted first.

[0091] Specifically, the voltage level and short-circuit capacity data are input into the weight coefficient table generation module, and each weight coefficient value is determined by table lookup method. For example, the 220kV voltage level corresponds to a straight line length weight coefficient of 0.6, and the 500kV corresponds to 0.8; the bending angle weight coefficient is 0.3 when the short-circuit capacity is 25kA, and it is increased to 0.5 when the short-circuit capacity is 50kA. After the bus segment cost value is weighted and summed, it enters the comparator with hysteresis characteristics, and the preset cost value threshold is set to the interval [1200, 1500]. When the cost value falls below 1200, the path output unit immediately generates an optimized path file; if it is in the interval of 1200-1500, the constraint adjustment unit is triggered to relax the constraint condition by 5%; when it exceeds 1500, the adjustment amplitude is increased to 10%. The constraint adjustment unit identifies the largest contribution factor by analyzing the proportion of each bus segment cost value component, for example, when the bending angle cost proportion exceeds 60%, the minimum bending radius constraint value is adjusted from 1.2D to 1.1D first. The adjusted constraint condition reenters the path generation process to form a closed-loop optimization until the cost value meets the threshold requirement.

[0092] As a preferred embodiment, the scheme of the present application is implemented as follows: when the topology total cost decision module is executed, first, the voltage level of the substation is 220 kV and the short-circuit capacity is 40 kA, and the weight coefficient table is automatically generated from the power system parameter library. In the weight coefficient table, the straight line length weight coefficient is 0.6, the bending angle weight coefficient is 0.3, and the support insulator quantity weight coefficient is 0.1. The bus section cost value output by each bus section calculation unit is input into the decision module, and each cost value is multiplied by the corresponding weight coefficient, for example, the bus section cost value is 85, and the contribution value after weighted calculation is 85x0.6=51. All the weighted contribution values are accumulated in the decision module to obtain the topology total cost of 320. The total cost is input into the digital comparator and compared with the preset threshold 300. Since 320 exceeds the threshold, the constraint adjustment unit triggers the automatic optimization mechanism: the minimum bending radius constraint is relaxed from 1.2 meters to 1.5 meters, the maximum allowable spacing of the support insulator is adjusted from 8 meters to 10 meters, and the grounding switch installation area boundary is expanded outward by 0.3 meters.

[0093] Through the above technical scheme, the present application effectively solves the problem of low efficiency and difficulty in achieving global optimization of manual adjustment of constraint conditions in the traditional method. The decision module automatically generates dynamic weight coefficients to balance the economy and safety of the path under different voltage levels; the closed-loop control mechanism of the comparator and the constraint adjustment unit makes the path optimization process adaptive and can automatically find the optimal solution under the premise of meeting engineering safety.

[0094] The above only describes the embodiments of the present application and is not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A substation insulated tubular busbar path routing optimization method, characterized by, The application relates to a parallelly-built insulating pipe type busbar topology mapping framework of a target transformer substation, and redundant nodes in equivalent networks of busbar sections are identified and removed; the insulating pipe type busbar topology mapping framework comprises a plurality of busbar section calculation units which can be independently operated, each busbar section calculation unit synchronously performs path optimization within a same clock beat; each busbar section calculation unit is driven by a comprehensive constraint set, takes starting connection point coordinates as input, iteratively generates busbar section discrete node coordinate chains, and quantitatively evaluates the busbar section discrete node coordinate chains through an embedded busbar section path evaluation model to obtain busbar section generation values, then all busbar section generation values are gathered to form a topology total generation cost; the busbar section path evaluation model takes a minimum bending radius criterion as a core and maps internal invisible nodes in a busbar section to external explicit nodes; a busbar level collaborative optimization framework of the target transformer substation is parallelly built; the busbar level collaborative optimization framework is sequentially connected by a constraint injection module, the insulating pipe type busbar topology mapping framework and a topology total generation cost decision module; the constraint injection module is used for issuing the comprehensive constraint set and starting connection point coordinates to the insulating pipe type busbar topology mapping framework; the topology total generation cost decision module is used for collecting and outputting the topology total generation cost of all busbar section calculation units; the parallelly-built insulating pipe type busbar topology mapping framework comprises the following steps: a busbar section equivalent network corresponding to each busbar section calculation unit is established according to a primary connection diagram of the target transformer substation; for each busbar section equivalent network, a corresponding busbar section path evaluation model is constructed; a node coordinate distributor and a generation value collector are configured for each busbar section calculation unit; the node coordinate distributor is used for solving and generating busbar section discrete node coordinate chains under the driving of the comprehensive constraint set and taking the starting connection point coordinates as a reference, and broadcasting the busbar section discrete node coordinate chains to all busbar section path evaluation models in the busbar section calculation unit; the generation value collector is used for collecting busbar section generation values output by all busbar section path evaluation models in the busbar section calculation unit and sending the results to the topology total generation cost decision module. the internal invisible nodes in the busbar section are mapped to the external explicit nodes by taking the minimum bending radius criterion as a core, which comprises the following steps:

2. The substation insulated tubular busbar path routing optimization method of claim 1, wherein, the minimum bending radius criterion is adopted to equivalently contract the busbar section discrete node coordinate chains in the busbar section path evaluation model, the internal invisible node coordinates are removed, and only the external explicit node coordinates are reserved; a busbar section generation value expression of the busbar section path evaluation model is as follows: E = alpha * L + beta * R + gamma * S in the formula, E represents the busbar section generation value, L represents a busbar section straight line length, R represents a busbar section bending angle, S represents a support insulator quantity, alpha, beta and gamma are respectively a straight line length weight coefficient, a bending angle weight coefficient and a support insulator weight coefficient; an external explicit node generation value expression after the equivalent contraction is as follows: Eext = alpha * Leq + beta * Req + gamma * Seq in the formula, Leq represents an equivalent straight line length, Req represents an equivalent bending angle, Seq represents an equivalent support insulator quantity, and Eext represents a busbar section generation value of the external explicit node. the corresponding busbar section path evaluation model is constructed, which comprises the following steps:

3. The substation insulated tub type bus duct path routing optimization method of claim 1, wherein, ​ The node decompression submodule, the internal state updating submodule, and the model compression submodule of the bus section path evaluation model are established; the node decompression submodule is used to read the bus section discrete node coordinate chain of the last iteration step, and restore all node coordinates in the equivalent network of the bus section; the internal state updating submodule is used to calculate the force history of the support insulator, the displacement history of the expansion joint, and the state history of the grounding switch in parallel; and the model compression submodule is used to aggregate the internal calculation results and output the bus section value.

4. The substation insulated tub type bus duct path routing optimization method of claim 1, wherein, The bus section equivalent network corresponding to each bus section calculation unit is established, including: Based on the Bezier curve control point mapping method, a path curve with bending radius constraint as the core in the bus section equivalent network is established; And / or, based on the principle of constant node coordinates, a straight line section equivalent path of the support insulator position in the bus section equivalent network is established.

5. The substation insulated tube bus path routing optimization method of claim 1, wherein, The method is realized based on a GIS platform or a BIM platform; When the method is realized based on the GIS platform, a multi-core CPU parallel architecture is used to realize hardware parallelism between different layers and between different cores in the same layer; When the method is realized based on the BIM platform, each module of the insulating tube type bus topology mapping framework and the bus level collaborative optimization framework is packaged into a sub-function, and a top-level scheduling function is created as an interactive interface of the BIM program and external interfaces; the top-level scheduling function is directly connected to all external interfaces, coordinates the data flow of each module, and controls the execution order of each module.

6. The substation insulated tube bus path routing optimization method of claim 1, wherein, The comprehensive constraint set includes a minimum bending radius constraint, a support insulator spacing constraint, and a grounding switch position constraint; The generation step of the minimum bending radius constraint includes reading the insulating tube type bus specification database, extracting the minimum allowable bending radius of the corresponding model, and writing into the global constraint register; the bus section path evaluation model calls the global constraint register every time a new node coordinate is generated, and if the new node coordinate causes the path section bending radius to be less than the minimum allowable bending radius, it is rolled back to the last node and the candidate node coordinate is regenerated; The generation step of the support insulator spacing constraint includes reading the substation civil structure database and extracting the full list of embedded part coordinates; when generating the bus section discrete node coordinate chain, if the spacing between adjacent nodes is greater than the maximum allowable spacing, a new node coordinate is inserted in the middle and forced to align with the nearest embedded part coordinate; The generation step of the grounding switch position constraint includes reading the primary equipment arrangement database and extracting the grounding switch installation area boundary list; when generating the bus section discrete node coordinate chain, if the node coordinate is outside any installation area boundary, the node coordinate is removed and a candidate node coordinate is regenerated.

7. The substation insulated tube bus path routing optimization method of claim 1, wherein, The generation of the bus section discrete node coordinate chain includes: Taking the minimum bending radius constraint, the support insulator spacing constraint, and the grounding switch position constraint as input conditions; Taking the starting connection point coordinate as the starting point, an improved A-star algorithm is used to calculate the initial bus section discrete node coordinate chain, wherein the heuristic function is the sum of the Euclidean distance and the constraint penalty term; Three times of spline smoothing are performed on the initial bus section discrete node coordinate chain to obtain the smoothed bus section discrete node coordinate chain; Read the three-dimensional collision model of the substation, and perform voxel collision detection on the node coordinates. If the node coordinates intersect with the obstacles, mark the conflict nodes. After removing the conflict nodes, call the improved A-star algorithm to regenerate the compensation node coordinates until the collision-free bus segment discrete node coordinate chain is obtained. Broadcast the collision-free bus segment discrete node coordinate chain to the bus segment path evaluation model.

8. The substation insulated tube bus path routing optimization method of claim 1, wherein, The calculation of the bus segment value includes: Read the insulating tube type bus material database, extract the unit length weight attribute, and calculate the bus segment linear length cost. Read the bending loss coefficient table, extract the unit angle bending loss, and calculate the bus segment bending angle cost. Read the support insulator cost table, extract the installation cost of a single support insulator, count the number of support insulators, and calculate the installation cost. Input the linear length cost, bending angle cost, and installation cost into the dynamic weight unit. The dynamic weight unit automatically outputs the weight coefficient based on the substation voltage level. After weighted summation, the bus segment value is obtained.

9. The substation insulated tube bus path routing optimization method of claim 1, wherein, The identification and removal of redundant nodes include: Traverse each node coordinate in the bus segment equivalent network, calculate the included angle value formed by the node coordinate and the previous and next node coordinates, and convert the included angle value to a curvature value. Compare the curvature value with the preset curvature threshold. If the curvature value is less than the preset curvature threshold, mark the node coordinate as a redundant node. Delete the redundant node coordinate, update the connection relationship of the previous and next node coordinates, and write the node coordinate into the node linked list of the bus segment equivalent network.

10. The substation insulated tub bus path routing optimization method according to any one of claims 1 to 9, characterized in that, The output of the topology total cost decision module includes: Read the substation voltage level and short-circuit capacity to automatically generate a weight coefficient table. Input the bus segment value of each bus segment calculation unit into the weight coefficient table to obtain the corresponding weight coefficient. Multiply the bus segment value by the weight coefficient and accumulate to obtain the topology total cost. Input the topology total cost into the comparator and compare it with the preset value threshold. If the topology total cost is less than the preset value threshold, call the path output unit to write the current bus segment discrete node coordinate chain into the optimization path file. If the topology total cost is not less than the preset value threshold, call the constraint adjustment unit to automatically relax the minimum bending radius constraint, increase the support insulator spacing, or expand the grounding switch installation area based on the exceeding amplitude. Return to the constraint condition generation step, bus segment discrete node coordinate chain generation step, bus segment value calculation step, and redundant node identification step until the topology total cost is less than the preset value threshold.

Images