Power transmission line operation and maintenance investment optimization method based on risk prediction

Abstract

The invention relates to the technical field of operation and maintenance investment optimization, in particular to a power transmission line operation and maintenance investment optimization method based on risk prediction, and the method comprises the steps: obtaining a meteorological disaster risk prediction result of each power transmission line section; a multi-dimensional operation and maintenance resource library is constructed and maintained, and a resource-risk benefit mapping relation representing that after execution of the resources in the library, specific risk quantification can be achieved, and benefits are reduced is established; integrating the risk prediction result, the resource attribute information and the mapping relation to construct a multi-constraint optimization model which takes maximization of the overall risk reduction benefit as a target function and takes total cost budget, resource availability constraint, risk control threshold and the like as constraint conditions; and solving the model to obtain an optimal operation and maintenance resource allocation scheme which specifies the type and quantity of the operation and maintenance resources to be allocated to each section, and outputting the optimal operation and maintenance resource allocation scheme. The defect that in the prior art, simple proportion distribution is carried out only according to the risk value, and resource cost differences and actual constraints are neglected is overcome.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of operation and maintenance investment optimization, and particularly relates to a power transmission line operation and maintenance investment optimization method based on risk prediction. BACKGROUND

[0002] In the field of power transmission line operation and maintenance management, meteorological disasters are one of the important factors leading to line faults and threatening the safe and stable operation of the power grid. In order to achieve precise allocation of operation and maintenance resources, the industry is generally committed to researching operation and maintenance investment optimization methods based on risk prediction. However, most of the existing methods often have the problem of single optimization dimension when converting risk prediction results into specific investment decisions. Specifically, these methods usually only rely on the predicted risk level or failure probability to make a simple proportional allocation of investment scale, without considering the differences in cost, availability and implementation benefits of multiple types of resources (such as manpower, materials, equipment, working hours, etc.) involved in operation and maintenance actions. This extensive allocation mode is difficult to achieve cost-effectiveness optimization under the constraint of limited operation and maintenance budget, and may lead to resource mismatch or low investment efficiency, which cannot meet the actual needs of fine and economic operation of power grid enterprises.

[0003] To solve the above problems, the existing technology such as Chinese invention patent CN114021844A power transmission line operation and maintenance investment optimization method and system based on meteorological disaster prediction divides meteorological, equipment and environmental three types of disaster-causing factors, fuses entropy weight method, analytic hierarchy process, deep self-encoding network and support vector machine, and constructs a composite fault prediction model to realize quantitative prediction of power grid faults caused by meteorological disasters. On this basis, the method calculates the operation and maintenance investment optimization coefficient according to the fault prediction result cumulative value of different regions, and then guides the allocation of operation and maintenance resources. This scheme effectively improves the accuracy of fault prediction and provides a preliminary quantitative basis for operation and maintenance investment decision-making, to a certain extent, improving the deficiency of relying on experience judgment.

[0004] However, the operation and maintenance investment optimization model of the above-mentioned disclosed method is essentially still a distribution strategy based on linear mapping of risk prediction results, and cannot break through the traditional framework. Specifically, the model only uses the cumulative value of the fault prediction result as the only input variable for optimization allocation, without further decomposing the operation and maintenance investment itself into multiple types of resource elements with different costs and different benefits, and without establishing a correlation model between investment cost and risk reduction benefit. Therefore, this method cannot achieve targeted, cost-effective multi-resource collaborative allocation scheme under the constraints of budget, manpower, materials, etc., and the fineness and economy of its optimization decision still need to be improved. SUMMARY

[0005] The present application aims to provide a power transmission line operation and maintenance investment optimization method based on risk prediction, which solves the problem that in the prior art, only simple linear distribution is performed according to the risk prediction value, and the cost difference and actual constraints of multiple types of operation and maintenance resources are not considered, so that the fine resource allocation decision with optimal cost benefit under the condition of limited budget cannot be realized.

[0006] To achieve the above-mentioned purpose, the present application provides a power transmission line operation and maintenance investment optimization method based on risk prediction, comprising the following steps:

[0007] Obtaining meteorological disaster risk prediction results of each power transmission line section in a target power grid region within a future preset period;

[0008] Building and maintaining a multi-dimensional operation and maintenance resource library, wherein the operation and maintenance resource attribute information of multiple operation and maintenance resources is defined, and the operation and maintenance resource attribute information includes resource type, cost attribute, availability constraint and efficiency attribute;

[0009] For part of the operation and maintenance resources in the operation and maintenance resource library, a resource-risk benefit mapping relationship is established, and the resource-risk benefit mapping relationship represents the reduction benefit value of a specific type of risk that can be achieved by executing a specific type or quantity of operation and maintenance resources;

[0010] Based on the meteorological disaster risk prediction results, the operation and maintenance resource attribute information and the resource-risk benefit mapping relationship, a multi-constraint optimization model is constructed, wherein the multi-constraint optimization model takes maximizing the overall risk reduction benefit of the target power grid region as the objective function, and at least one of the total cost budget, the availability constraint of each type of operation and maintenance resource, and the risk control threshold of the specified line section as the constraint condition;

[0011] Solving the multi-constraint optimization model to obtain an optimal operation and maintenance resource allocation scheme, wherein the optimal operation and maintenance resource allocation scheme specifies the type and quantity of operation and maintenance resources that should be allocated to each power transmission line section;

[0012] Outputting the optimal operation and maintenance resource allocation scheme.

[0013] Among them, the building and maintaining a multi-dimensional operation and maintenance resource library specifically includes:

[0014] Dividing the operation and maintenance resources into at least two of human patrol resources, special operation vehicle resources, unmanned aerial vehicle resources, material spare parts resources and online monitoring device resources, and defining the corresponding unit cost, regional inventory quantity, maximum schedulable quantity and single operation standard efficiency parameter for each resource.

[0015] Among them, the establishing a resource-risk benefit mapping relationship specifically includes:

[0016] Based on historical operation and maintenance data or simulation models, at least one of the dancing risk caused by strong wind, the lightning trip-out risk caused by lightning or the mechanical overload risk caused by icing is determined, and at least one recommended operation and maintenance resource and its estimated risk reduction benefit value corresponding to each risk are determined.

[0017] Among them, based on the meteorological disaster risk prediction result, the operation and maintenance resource attribute information and the resource-risk benefit mapping relationship, a multi-constraint optimization model is constructed, specifically including:

[0018] A linear programming model or integer programming model is constructed, which maximizes the overall risk reduction benefit while meeting the upper limit of the total cost budget, the upper limit of the available number of each type of operation and maintenance resource, and the risk value of the key power transmission channel being lower than the preset threshold.

[0019] Among them, the method further includes:

[0020] The process of obtaining risk prediction results, constructing and solving the multi-constraint optimization model is executed in a preset time period, and the latest real-time state information of power grid equipment and real-time inventory information of operation and maintenance resource library are updated as part of the constraint conditions each time the solution is solved.

[0021] Among them, the solving of the multi-constraint optimization model specifically includes:

[0022] By adjusting the numerical value of the total cost budget, a plurality of comparative operation and maintenance resource allocation schemes corresponding to high, medium and low different budget levels are solved respectively, and the expected overall risk reduction rate and resource utilization rate corresponding to each comparative scheme are calculated.

[0023] Among them, the output of the optimal operation and maintenance resource allocation scheme specifically includes:

[0024] A decision support report is generated, which compares and displays the plurality of comparative operation and maintenance resource allocation schemes in a visual form, and presents the resource flow distribution, the expected risk reduction effect comparison and the cost-benefit analysis under each scheme.

[0025] The power transmission line operation and maintenance investment optimization method based on risk prediction provided by the present application comprises the following steps: BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced.

[0027] Figure 1 is a step flow chart of the power transmission line operation and maintenance investment optimization method based on risk prediction of the present application. DETAILED DESCRIPTION

[0028] The embodiments of the present application will be described in detail below, and the examples of the embodiments are shown in the drawings. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0029] Please refer to Figure 1 The present application provides a power transmission line operation and maintenance investment optimization method based on risk prediction, which comprises the following steps:

[0030] S101: Obtain the meteorological disaster risk prediction results of each power transmission line section in a target power grid region within a future preset period;

[0031] Specifically, first, the range of the target power grid region is determined, and the power transmission lines in the region are divided into several independent line sections according to the power grid topology, geographical environmental characteristics or operation and maintenance management habits. Each section can be a specific line corridor, such as a line between two specific towers. The purpose of the division is to enable subsequent differentiated risk assessment and resource allocation.

[0032] Secondly, obtain the meteorological disaster risk prediction result of each line section in a future preset period (for example, 24 hours, 72 hours or a week in the future). The risk prediction result is a quantitative value for representing the probability of failure or the expected loss degree of the section caused by a specific meteorological disaster (such as strong wind, lightning, hail, icing, etc.) in the preset period. The result can be obtained by relying on existing and mature meteorological disaster disaster-causing and power grid failure prediction models.

[0033] As a realizable preferred mode, the composite failure prediction model disclosed in the Chinese invention patent CN114021844A mentioned in the background can be directly quoted or accessed. The model can output a prediction value representing the probability of power grid failure by fusing meteorological factors, equipment factors and environmental factors, and using trained neural network and support vector machine model. The real-time and forecast meteorological data, equipment account and state data, geographical environment data, etc. corresponding to each line section after the target area is divided are submitted as input data batches to the composite failure prediction model. After the model is processed, one or more failure risk prediction values corresponding to different meteorological disaster types are output for each line section. For example, for a certain section, the model may output a "lightning trip-out risk value of 0.15" and a "strong wind dancing risk value of 0.08".

[0034] Further, in order to adapt to the needs of the subsequent optimization model for unified risk measurement, it can be necessary to standardize or normalize the original prediction results obtained. For example, the probability values output by different models or indicators of different dimensions are uniformly mapped to a preset risk value interval, such as between 0 and 1 or between 0 and 100, by linear transformation or other methods, to form the "meteorological disaster risk prediction result" referred to in the present application. The result is a structured data set that clearly records the identifier of each line section and its corresponding quantified risk value.

[0035] The execution can be periodic (such as automatically executed once every 6 hours) or triggered by a specific event (such as the issuance of a disaster weather warning). The macro meteorological forecast is converted into fine-grained risk information that can be quantitatively compared and linked to specific power grid assets, laying a reliable data foundation for subsequent cost-effective precise resource optimization allocation.

[0036] S102: build and maintain a multi-dimensional operation and maintenance resource library, wherein a plurality of operation and maintenance resource attribute information of various operation and maintenance resources are defined, and the operation and maintenance resource attribute information includes resource type, cost attribute, availability constraint and efficiency attribute;

[0037] Specifically, first, all tangible and intangible inputs involved in the transmission line operation and maintenance work are systematically sorted and classified. According to the role, form and cost characteristics of resources in operation and maintenance activities, they can be divided into several categories. For example, it can include but not limited to: human inspection resources (such as different skill levels of line inspectors, maintenance teams), special operation vehicle resources (such as aerial work platforms, live-line work vehicles, engineering rescue vehicles), unmanned aerial vehicle inspection resources (including unmanned aerial vehicle platforms, control personnel and data processing units), material spare parts resources (such as insulators, conductors, fittings, lightning arresters, fuses) and online monitoring device resources (such as micro meteorological stations, image and video monitoring devices, conductor temperature sensors). Each resource in the library is defined as an independent data object.

[0038] Secondly, the multi-dimensional "operation and maintenance resource attribute information" of each resource in the resource library is defined, which is the basis for mathematical modeling and optimization calculation. Key attributes include:

[0039] Resource type: As mentioned above, the classification identifier is used to distinguish the fundamental nature of the resource.

[0040] Cost attribute: refers to the economic cost of using or consuming the resource. Usually includes:

[0041] Unit purchase / rental cost: the cost of obtaining the resource itself.

[0042] Unit transportation / distribution cost: the cost of transporting the resource to the work site.

[0043] Unit operation time cost: the cost of using resources for operation, such as labor, energy consumption, etc. For example, for a line inspection team, its cost attribute may include daily wage total, vehicle daily oil consumption and wear and tear conversion cost.

[0044] Availability constraints: refers to the number limit of the resource that can be called within the optimization period. Usually includes:

[0045] Regional inventory quantity: the physical quantity currently stored in regional warehouses or bases.

[0046] Maximum schedulable quantity: considering cross-regional support, rental channels and other factors, the maximum quantity that can actually be put into use within the optimization period.

[0047] Replenishment period: for consumable materials, the time required from application to arrival on site, which affects its availability in rolling optimization.

[0048] Efficiency attribute: refers to the technical effect or coverage that the resource can produce for a standard operation. Usually includes:

[0049] Single operation coverage mileage / tower number: applicable to inspection resources.

[0050] Standard operation time: the man-hour required to complete a typical operation (e.g. replacing a set of insulators).

[0051] Risk reduction performance coefficient: this attribute is closely related to the subsequent "resource-risk benefit mapping relationship" and can be used as a basic parameter. For example, the performance attribute of a certain type of lightning arrester may include the basic data of "expected reduction of single-pole tower lightning trip-out probability X%".

[0052] Then, the collection and dynamic maintenance mechanism of the resource library data is established. The initial values of the above attribute information can be obtained and integrated from the existing multiple information systems of the enterprise, for example:

[0053] Resource type, model, unit purchase cost, etc. Information comes from material management system or asset management system.

[0054] Regional inventory quantity, location, etc. Information comes from warehouse management system or ERP system.

[0055] Human cost, team scheduling plan comes from human resource management system or production management system.

[0056] The available state and operation performance parameters of special vehicles and unmanned aerial vehicles may come from the equipment management system or the historical operation database.

[0057] These data are synchronized to the "multi-dimensional operation and maintenance resource library" of the present application through a pre-set data interface or ETL process, on a regular basis (such as daily) or triggered, to ensure that the resource information used by the optimization model is consistent with the actual situation. The resource library can be logically represented as a relational database table or a series of structured data files, with each record corresponding to a specific resource instance and the fields corresponding to the above attributes.

[0058] The operation and maintenance management is sunk from the traditional "funding budget allocation" level to the specific "resource entity scheduling" level, so that the subsequent optimization model can operate under clear cost, quantity and capacity constraints, laying a solid data foundation for realizing fine and cost-effective decision-making.

[0059] S103: For part of the operation and maintenance resources in the operation and maintenance resource library, a resource-risk benefit mapping relationship is established, which represents the reduction benefit value of a specific type or quantity of operation and maintenance resources to a specific type of risk;

[0060] Specifically, first, the data structure of the "resource-risk benefit mapping relationship" is determined. This relationship is essentially a many-to-many association rule base or parameter matrix. Each mapping relationship contains at least the following core elements: target risk type (such as "strong wind dancing risk", "lightning strike trip risk"), applicable operation and maintenance resource type and quantity (such as "deploy 1 unmanned aerial vehicle inspection team, work for 2 hours", "install 3 sets of phase-to-phase spacer bars"), and corresponding estimated risk reduction benefit value (a quantitative value, such as reducing the corresponding risk value of a specific section by 0.05). The benefit value needs to use the same dimension and scale as the risk prediction results obtained in step S101 to ensure that it can be directly used for calculation.

[0061] Second, the main data sources and construction methods for establishing the mapping relationship include the following two types:

[0062] Statistical analysis based on historical operation and maintenance data: This is the most direct and reliable method to establish the mapping relationship. In specific operation, historical data of the same period is extracted from the production management system (PMS), equipment management system, fault recording system, and material management system of the power grid enterprise. For example: for a line section that has experienced strong wind dancing fault, query its operation and maintenance records in a certain period before the fault, such as whether unmanned aerial vehicle special patrol has been conducted, whether wind-resistant guy has been installed or reinforced, etc.

[0063] Through data cleaning and correlation analysis, the fault occurrence rate or risk index change under similar meteorological events is compared between the area that has taken a certain specific operation and maintenance measure and the similar area that has not taken the measure.

[0064] Using statistical regression methods (such as multiple linear regression, logistic regression), the contribution of different operation and maintenance measures to risk reduction is calculated, and the approximate benefit value is obtained. For example, analysis may show that: "in icing prone sections, spraying anti-icing paint in advance can reduce the icing trip risk prediction value of the section by an average of about 30%".

[0065] Derivation calculation based on physical model or simulation model: For new resources or new risk scenarios that lack sufficient historical data, simulation methods can be used. For example:

[0066] For "lightning risk", electromagnetic transient simulation software (such as EMTP / ATP) can be used to establish models of transmission lines and towers, and to simulate and calculate the change of insulator string voltage caused by lightning current before and after installing different types of lightning arresters, thereby quantitatively evaluating the effect of reducing lightning flashover probability (i.e. risk).

[0067] For "dancing risk", computational fluid dynamics (CFD) simulation can be used to simulate the improvement of conductor aerodynamic characteristics under different spacer bar installation schemes, and then convert it into the reduction of dancing amplitude and the corresponding risk reduction value.

[0068] The results of these simulations, after being verified against a small amount of field data, can be converted into entries in a mapping table.

[0069] Then, the mapping relationships obtained from the above analysis or simulation are structured and stored to form a "resource-risk-benefit knowledge base" that can be queried and called by the optimization model. This knowledge base can be a database table, whose fields include at least: risk type code, resource type code, resource quantity, benefit value, and confidence level (reflecting the data support strength of the relationship). For example, a record might be: Risk type = "lightning trip", Resource type = "line-type surge arrester", Unit quantity = "1 set installed per tower", Estimated benefit value = "reduces the risk value of this tower by 0.08".

[0070] Finally, this mapping database is a living, updatable knowledge system. With the continuous accumulation of operational data, the application of new technologies, and changes in the power grid structure, the mapping relationships are reviewed and revised periodically (e.g., quarterly or annually). For example, when a new type of online monitoring device is put into use, it is necessary to analyze the number of faults avoided through real-time early warnings using operational data over a period of time, thereby supplementing or updating its mapping relationships with "wildfire risk," "external damage risk," etc.

[0071] This process solidifies expert experience, historical patterns, and simulation conclusions into machine-readable and computable quantitative rules. This enables subsequent optimization models to move beyond blind resource allocation and instead make intelligent decisions based on scientific predictions of "input-output" benefits. It fundamentally addresses the deficiency of lacking a model linking input costs and risk reduction benefits, providing crucial core logical support for achieving cost-effective and refined resource allocation.

[0072] S104: Based on the meteorological disaster risk prediction results, the operation and maintenance resource attribute information, and the resource-risk benefit mapping relationship, a multi-constraint optimization model is constructed. The multi-constraint optimization model takes maximizing the overall risk reduction benefit of the target power grid area as the objective function, and takes at least one of the following as constraints: total cost budget, availability constraints of various operation and maintenance resources, and risk control threshold of a specified line section.

[0073] Specifically, define the decision variables of the model. These are the unknowns that the model needs to solve for, directly corresponding to the final resource allocation scheme. Typically, a set of decision variables can be defined. ,in Index representing a section of a transmission line ( ), Indexes representing operation and maintenance resource types ( ).variable Representative assigned to the first the first number of operation and maintenance resources (or whether to perform a certain job, represented by 0 or 1) of each line section. These variables constitute a resource allocation matrix, the solution of which will directly guide the operation and maintenance actions.

[0074] Secondly, the objective function of the model is constructed. The core optimization goal of the present application is to maximize the "overall risk reduction benefit" of the entire target power grid area. This benefit is the sum of the risk reduction values achieved by each section due to the allocation of operation and maintenance resources. Specifically:

[0075] For a certain section , its initial risk value is given by the meteorological disaster risk prediction result of step S101.

[0076] If it is allocated a resource combination (determined by decision variables ), according to the "resource-risk benefit mapping relationship" established in step S103, the risk reduction benefit value of each type of risk of this section for this resource combination can be calculated.

[0077] The final risk reduction benefit of this section can usually be expressed as the sum of the risk reduction values brought by all the resources allocated to it, i.e. .

[0078] Therefore, the objective function of the entire area can be formally defined as maximizing the total benefit: . The key here is the specific form of the function, which encapsulates the mapping knowledge in step S103, which can be a linear function (for example, installing a set of lightning arresters to fixedly reduce a certain amount of risk), or a more complex nonlinear relationship.

[0079] Then, the constraints of the model are established. This is the key to integrating real-world restrictions into the mathematical model of the present application, and is also the core that distinguishes it from simple proportional allocation. The main constraints include:

[0080] Total cost budget constraint: the total cost of all allocated resources cannot exceed a preset upper limit . The unit cost of a resource comes from the "cost attribute" of step S102. The constraint expression is: .

[0081] Resource availability constraint: the total amount of each resource allocated cannot exceed its current available quantity . The "availability constraints" from step S102 (such as regional inventory and maximum schedulable quantity) are expressed as follows: .

[0082] Risk control threshold constraints for critical sections: For certain particularly important transmission channels (such as inter-regional trunk lines and lines supplying power to important users), it may be required that their ultimate risk be controlled within a certain threshold. The following is a section. If a segment belongs to this set of key segments, then the constraint is: This ensures that the optimization scheme, while pursuing overall benefits, also meets the mandatory local security requirements.

[0083] Nonnegativity and Integer Constraints: Decision variables are typically required to be nonnegative integers, i.e. This aligns with the physical reality that resources are indivisible.

[0084] Combining the above objective function and constraints, a typical constrained optimization problem is thus constructed. According to... Depending on the properties of the function and constraints, the model may be concretized as a linear programming (LP) model (if the benefit function and all constraints are linear), an integer linear programming (ILP) model (requiring variables to be integers based on LP), or a more general nonlinear programming model. These are all mature model frameworks in the field of operations research, and their construction process is clear to those skilled in the art.

[0085] This invention successfully abstracts and transforms a complex operation and maintenance management decision-making problem into a well-structured and precisely defined mathematical optimization problem. The model integrates multi-source heterogeneous information, including risk information, resource entities, cost data, and business rules (thresholds), providing a precise mathematical foundation for finding globally optimal or near-optimal allocation schemes through automated calculation under strict real-world constraints. This marks a fundamental shift from an experience-based, extensive decision-making model to a data- and model-based, refined, and scientific decision-making model.

[0086] S105: Solve the multi-constraint optimization model to obtain the optimal operation and maintenance resource allocation scheme, wherein the optimal operation and maintenance resource allocation scheme specifies the type and quantity of operation and maintenance resources to be allocated to each transmission line section;

[0087] Specifically, firstly, based on the specific model type determined in step S104, the corresponding optimization algorithm and tools are selected and configured. This constitutes the technical basis of the solution process:

[0088] If the model is constructed as a linear programming (LP) model, that is, the objective function and all constraints are decision variables. linear expression, then mature linear programming solvers can be used to solve it. Such solvers are usually based on classical algorithms such as Simplex Method or Interior Point Method, and can efficiently and reliably find the global optimal solution. For example, one can call the LP solvers integrated in Gurobi, CPLEX, MATLAB’s linprog function, or Python’s PuLP, SciPy.optimize, etc.

[0089] If the model is constructed as an integer linear programming (ILP) or mixed integer linear programming (MILP) model, i.e., some or all of the decision variables are required to take integer values (such as the number of devices, the number of work teams), then solvers suitable for integer programming need to be used. Such solvers usually integrate algorithms such as Branch and Bound, Cutting Plane, etc. to handle the discreteness of variables. Similarly, commercial solvers such as Gurobi, CPLEX, or open-source libraries such as OR-Tools provide strong integer programming solving capabilities.

[0090] If the model involves nonlinear relationships, then nonlinear programming (NLP) solvers may need to be called to solve it using algorithms such as sequential linear programming, sequential quadratic programming, or gradient descent. In practice, by designing a reasonable mapping relationship, it is often preferred to construct the problem as a linear or integer linear model to take advantage of its maturity and efficiency in solving.

[0091] Secondly, model instantiation and solving calculation are performed. This process inputs the specific data prepared in the previous steps into the solver:

[0092] Input model parameters: input the specific parameters of the risk value of each line segment obtained in step S101 , the unit cost and available quantity of each type of resource obtained in step S102 , the specific parameters (such as linear coefficients) of the resource-risk benefit function defined in step S103 , as well as the preset total budget and key segment risk threshold , etc. in numerical form into the solver, completing the instantiation of the abstract model. Perform solving operation: call the solver to solve the instantiated mathematical model above. The solver will automatically execute its core algorithm (such as iteration of the Simplex table, search of the Branch and Bound tree) to find the solution that satisfies all the constraints (total cost budget constraint, resource availability constraint, risk threshold constraint, etc.) and can make the objective function

[0093] ​A set of optimal numerical solutions of the decision variables .

[0094] Processing and verifying the solution: after the solution is completed, the status of the solution output by the solver (such as "optimal", "feasible" or "no solution") and the optimal value of the decision variable are obtained . It is usually necessary to verify the solution, such as checking whether all constraints are strictly satisfied, and the rationality of the solution in business logic (such as whether the allocation quantity is a non-negative integer).

[0095] Then, the solution result is explained, and an optimal operation and maintenance resource allocation scheme is generated. The numerical matrix of the decision variable is abstract and needs to be converted into a scheme that business personnel can understand and execute:

[0096] Scheme structuring: convert the optimal solution into a structured "operation and maintenance resource allocation schedule". The schedule takes the line section as the row and the operation and maintenance resource type as the column, and the numerical value in each cell of the table is the number of a specific resource that should be allocated to the section according to the solution. For example, indicates that 2 units of the 5th type of resource (such as 2 drones) should be allocated to the 3rd line section.

[0097] Detailed description of the scheme: based on the schedule, a detailed scheme description can be further generated. It includes: a list of resources to be allocated to each section and the number of resources, the expected risk reduction value (calculated by ), the cost of allocating resources to the section (calculated by ), and the overall risk reduction benefit and total cost after aggregation.

[0098] Scheme executability conversion: for resources such as "human patrol" and "special operation" that require specific operation time and path, the allocated number needs to be further refined into specific operation task instructions in combination with the standard operation parameters of the resource (such as single operation coverage mileage, standard working hours), such as "dispatch patrol team A to conduct special patrol on the 5th-10th tower section of XX line from 9:00 to 11:00 tomorrow".

[0099] The present application has completed the key leap from data, knowledge and model to final decision. The solution process fully utilizes the mature and reliable computing tools in the field of operations research, ensuring the mathematical optimality of the decision scheme. The "optimal operation and maintenance resource allocation scheme" generated finally is a specific, quantitative and directly actionable plan that can guide operation and maintenance activities, which clearly answers the core questions of "where, what and how much", so that the theoretical optimization model constructed in step S104 finally becomes a highly operational management instruction. This is the final computing link to realize fine and scientific operation and maintenance resource scheduling.

[0100] S106: output the optimal operation and maintenance resource allocation scheme.

[0101] Specifically, first, different forms and granularities of output content are generated according to the requirements of application scenarios and decision levels. The main output includes:

[0102] Structured operation and maintenance resource allocation plan table: the optimal value matrix of decision variables obtained in step S105 is automatically filled into a predefined electronic form or database table. The table takes the transmission line section as the row and the operation and maintenance resource type as the column, clearly showing the specific number of each type of resource that each section should obtain. For example, a row of data in the table can display: "Section ID: LS-101, allocation: 2 unmanned aerial vehicle inspections, 3 lightning arresters, 1 special patrol team". This plan table is the direct basis for all subsequent scheduling and execution work.

[0103] Executable task order and scheduling instructions: for resources that require on-site work (such as manpower, vehicles), the system further decomposes and generates specific, executable electronic work orders that can be issued to teams or individuals, combining abstract allocation quantities with information such as "standard operation time" and "single coverage range" in resource attributes. For example, based on the decision "allocate 1 special patrol team to section LS-101", the system can automatically generate a work order containing: work team name, work leader, work section start and end tower numbers, planned work time, required equipment list, safety precautions, etc. At the same time, scheduling instructions for vehicles and large machinery can be generated, clearly indicating the calling object, departure location, destination, and time window.

[0104] Visual decision support report: this is a high-level output for management decision makers. An illustrated report is automatically generated, with core content including:

[0105] Resource flow heat map: on a geographic information system (GIS) map, the allocation and aggregation of manpower, materials, and other resources from warehouses / bases to each high-risk line section are visually displayed with different colors and arrows.

[0106] Risk reduction effect comparison chart: through bar charts or curve charts, the decline in predicted risk values for each region and the overall risk after implementing the optimization scheme compared to the original risk values is displayed, clearly presenting the "safety benefits" of the investment.

[0107] Multi-scenario scheme comparison and analysis: if multiple budget levels have been optimized and solved, the report will display multiple schemes under "high, medium, and low" different budgets side by side, and compare their expected overall risk reduction rates, total costs, resource utilization rates, and other key performance indicators (KPIs), providing decision makers with the basis for weighing options. ​

[0108] Cost-benefit analysis summary: list the total budget, the cost proportion of each major resource category, the average risk reduction value brought by unit cost, etc. in table form, reflecting the economic nature of the scheme.

[0109] Secondly, establish diversified output channels and interfaces to ensure smooth delivery of the scheme to different users and systems:

[0110] Human-machine interface (HMI) display: through web pages or client applications, the above schedule, work order and visual report are pushed to relevant management personnel of power grid dispatching center and operation and maintenance management department in real time. The interface supports interactive operations such as filtering, drilling and downloading.

[0111] Integration with production management system (PMS): through standardized application programming interface (API), the generated work order and resource demand list are automatically synchronized or written into the existing production management system of the enterprise, directly entering the work order circulation and execution tracking process, realizing seamless connection with the existing operation system.

[0112] Mobile terminal push: specific operation work orders and safety prompts are immediately pushed to the smart terminals of front-line operation personnel through mobile applications or short messages, guiding on-site operation.

[0113] Generation of standardized files: automatically generate standardized files such as PDF version of decision report and Excel format of resource allocation schedule table for archiving, reporting or further distribution.

[0114] Finally, the output process should have confirmation and feedback mechanisms. For example, relevant management personnel can be required to confirm the main scheme online; front-line personnel can feed back the receiving status after receiving the work order; the resource warehouse can feed back the delivery status after receiving the material allocation instruction. These feedback information can be fed back to the system for resource state update in subsequent rolling optimization.

[0115] The present application completes the complete closed loop from data collection, model construction, optimization solution to achievement delivery. It not only produces a mathematical optimal solution, but also through a series of output forms close to business practice, converts the optimal solution into specific, clear and traceable instructions and information driving actual operation actions, truly realizes the empowerment of intelligent decision analysis capability in daily operation and management business, and solves the pain points of abstract decision results and poor executability of traditional methods.

[0116] The above only discloses one or more preferred embodiments of the present application, which cannot limit the scope of the rights of the present application. Those skilled in the art can understand that all or part of the processes of the above embodiments are implemented, and equivalent changes made according to the claims of the present application still belong to the scope covered by the present application.

Claims

1. A method for optimizing the operation and maintenance investment of transmission lines based on risk prediction, characterized in that, Includes the following steps: Obtain meteorological disaster risk prediction results for each transmission line section within the target power grid area within a preset future period; Construct and maintain a multi-dimensional operation and maintenance resource library, which defines the operation and maintenance resource attribute information of various operation and maintenance resources, including resource type, cost attribute, availability constraint and performance attribute; For some of the operation and maintenance resources in the operation and maintenance resource library, a resource-risk-benefit mapping relationship is established. The resource-risk-benefit mapping relationship represents the reduction benefit value of a specific type of risk that can be achieved by executing a specific type or number of operation and maintenance resources. Based on the meteorological disaster risk prediction results, the operation and maintenance resource attribute information, and the resource-risk benefit mapping relationship, a multi-constraint optimization model is constructed. The multi-constraint optimization model takes maximizing the overall risk reduction benefit of the target power grid area as the objective function, and takes at least one of the following constraints: total cost budget, availability constraints of various operation and maintenance resources, and risk control threshold of a specified line section. Solving the multi-constraint optimization model yields the optimal operation and maintenance resource allocation scheme, which specifies the type and quantity of operation and maintenance resources to be allocated to each transmission line section. Output the optimal operation and maintenance resource allocation scheme.

2. The method for optimizing transmission line operation and maintenance investment based on risk prediction as described in claim 1, characterized in that, The construction and maintenance of a multi-dimensional operation and maintenance resource library specifically includes: Operation and maintenance resources are divided into at least two of the following: human inspection resources, special operation vehicle resources, drone resources, material spare parts resources, and online monitoring device resources. For each resource, the corresponding unit cost, regional inventory quantity, maximum dispatchable quantity, and standard performance parameters for a single operation are defined.

3. The method for optimizing transmission line operation and maintenance investment based on risk prediction as described in claim 1, characterized in that, The establishment of the resource-risk-benefit mapping relationship specifically includes: Based on historical operation and maintenance data or simulation models, at least one recommended operation and maintenance resource and its estimated risk reduction benefit value are determined for at least one of the following risks: galloping risk caused by strong winds, lightning tripping risk caused by lightning, or mechanical overload risk caused by icing.

4. The method for optimizing transmission line operation and maintenance investment based on risk prediction as described in claim 1, characterized in that, Based on the meteorological disaster risk prediction results, the operation and maintenance resource attribute information, and the resource-risk-benefit mapping relationship, a multi-constraint optimization model is constructed, specifically including: Construct a linear programming model or integer programming model that aims to maximize the benefits of overall risk reduction while simultaneously satisfying the upper limit of total cost budget, the upper limit of the number of available operation and maintenance resources, and the risk value of key transmission channels being lower than a preset threshold.

5. The method for optimizing transmission line operation and maintenance investment based on risk prediction as described in claim 1, characterized in that, The method further includes: The process of obtaining risk prediction results, constructing and solving the multi-constraint optimization model is executed in a rolling manner at a preset time period, and the latest real-time status information of power grid equipment and the real-time inventory information of the operation and maintenance resource library are updated as part of the constraints during each solution.

6. The method for optimizing transmission line operation and maintenance investment based on risk prediction as described in claim 1, characterized in that, Solving the multi-constraint optimization model specifically includes: By adjusting the total cost budget, multiple comparative operation and maintenance resource allocation schemes corresponding to high, medium, and low budget levels are obtained, and the expected overall risk reduction rate and resource utilization rate of each comparative scheme are calculated.

7. The method for optimizing transmission line operation and maintenance investment based on risk prediction as described in claim 6, characterized in that, The output of the optimal operation and maintenance resource allocation scheme specifically includes: Generate a decision support report that visually compares and contrasts the multiple comparative operation and maintenance resource allocation schemes, and presents the resource flow distribution, expected risk reduction effect comparison, and cost-benefit analysis under each scheme.

Images