Cable branch box-based cost optimization method, device, equipment and storage medium

Abstract

The present application relates to the technical field of cost optimization, and discloses a cost optimization method and device based on a cable branch box, equipment, and a storage medium; equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault maintenance parameters are subjected to cost analysis to obtain predicted full-cycle costs and calculated full-cycle costs; the predicted full-cycle costs and the calculated full-cycle costs are subjected to redundancy analysis according to a preset redundancy analysis method to obtain equipment capacity redundancy rates; a cable branch box procurement table is obtained, and the cable branch box procurement table is subjected to cost optimization according to the equipment capacity redundancy rates and a preset cable branch box type library to obtain an optimized cost procurement table; five types of cost parameters are collected to account for predicted and actual full-cycle costs, the equipment capacity redundancy rates are obtained through redundancy analysis, and idle capacity is identified; the optimized procurement table is combined with the cable branch box type library to match the specifications, thereby realizing fine economic management and control of power distribution materials.

Description

Technical Field

[0001] This invention relates to the field of cost optimization technology, and in particular to cost optimization methods, apparatus, equipment and storage media based on cable branch boxes. Background Technology

[0002] In various power distribution construction scenarios, such as urban and rural power distribution networks, residential power distribution projects, and industrial and commercial power supply systems, cable branch boxes, as terminal power distribution equipment, directly determine the overall project investment through their selection, procurement, and full-cycle cost control. Currently, the industry generally relies on static rated load standards to complete box selection, and cost accounting simply calculates the unit price of purchase without taking into account the entire chain of expenditures, including transportation and installation, annual operation and maintenance, power loss, and fault compensation. At the same time, it ignores the time value of money, resulting in distorted calculation results. Existing procurement methods cannot combine time-series load fluctuations and raw material price cycles to dynamically assess the actual equipment needs, making it difficult to distinguish between seasonal load surpluses and long-term permanent capacity redundancy. Redundancy determination lacks quantitative indicators, which easily leads to the over-configuration of large-sized boxes. The optimization of existing procurement lists lacks standardized data support and relies solely on manual experience to adjust models, which can easily cause continuous ineffective costs in multiple stages of procurement, construction, and operation and maintenance. The cost control of power distribution materials is extensive and cannot meet the development needs of lean and digital cost control in power engineering. Summary of the Invention

[0003] In order to overcome the shortcomings of the prior art, the present invention aims to provide a cost optimization method, apparatus, equipment and storage medium based on cable branch boxes.

[0004] The cost optimization method based on cable branch boxes includes: collecting vehicle access requests and vehicle access distances; searching a preset vehicle permission database based on the vehicle access requests and vehicle access distances to obtain vehicle credential information, vehicle permissions, and remaining access counts; adjusting vehicle permissions based on preset access count thresholds, vehicle credential information, and remaining access counts to obtain vehicle access permissions; and enabling vehicles to enter the floor access control state based on preset corridor control mechanisms and vehicle access permissions.

[0005] Furthermore, the cost optimization method based on cable branch boxes includes: obtaining equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters; performing cost analysis on the equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain predicted full-cycle costs and calculated full-cycle costs; performing redundancy analysis on the predicted full-cycle costs and calculated full-cycle costs according to a preset redundancy analysis method to obtain the equipment capacity redundancy rate; obtaining the cable branch box procurement table, and optimizing the cost of the cable branch box procurement table according to the equipment capacity redundancy rate and a preset cable branch box type library to obtain an optimized cost procurement table.

[0006] Furthermore, the cost analysis of equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain predicted and calculated full-cycle costs includes: collecting procurement demand parameters, supplier datasets, equipment full-lifecycle datasets, and historical equipment cost data; pre-training a preset PSO-XGBoost regression model based on the procurement demand parameters, supplier datasets, equipment full-lifecycle datasets, historical equipment cost data, and a preset window aggregation method to obtain a cost prediction model; performing predictive analysis on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters based on the cost prediction model to obtain predicted full-cycle costs; and calculating equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters based on preset cost item calculation formulas and preset time value of money discount factors to obtain calculated full-cycle costs.

[0007] Furthermore, the step of pre-training a pre-defined PSO-XGBoost regression model based on procurement demand parameters, supplier datasets, equipment lifecycle datasets, historical equipment cost data, and a pre-defined window aggregation method to obtain a cost prediction model includes: performing feature analysis on procurement demand parameters, supplier datasets, and equipment lifecycle datasets based on a pre-defined feature analysis method to obtain lag features and market information features; performing aggregation statistics on lag features and market information features based on a pre-defined sliding time window and window aggregation method to obtain sliding statistical features; and pre-training the PSO-XGBoost regression model based on historical equipment cost data, sliding statistical features, lag features, and market information features to obtain a cost prediction model.

[0008] Furthermore, the step of pre-training the PSO-XGBoost regression model based on historical equipment cost data, sliding statistical characteristics, lag characteristics, and market information characteristics to obtain a cost prediction model includes: optimizing the parameters of the PSO-XGBoost regression model based on lag characteristics, sliding statistical characteristics, and market information characteristics to obtain an optimized regression model; and pre-training the optimized regression model based on historical equipment cost data to obtain a cost prediction model.

[0009] Furthermore, the step of performing redundancy analysis on the predicted full-cycle cost and the calculated full-cycle cost according to a preset redundancy analysis method to obtain the equipment capacity redundancy rate includes: determining whether the calculated full-cycle cost is greater than a preset cycle cost threshold; if the calculated full-cycle cost is greater than the cycle cost threshold, then obtaining the floor's electricity demand parameters; and performing redundancy analysis on the calculated full-cycle cost and the predicted full-cycle cost according to the redundancy analysis method and the floor's electricity demand parameters to obtain the equipment capacity redundancy rate.

[0010] Furthermore, the step of performing redundancy analysis on the calculated and predicted full-cycle costs based on redundancy analysis methods and floor electricity demand parameters to obtain the equipment capacity redundancy rate includes: predicting floor electricity demand parameters based on a preset improved LSTM time series prediction algorithm and a preset seasonal fluctuation coefficient to obtain the electricity load change trend characteristics; and performing redundancy analysis on the calculated and predicted full-cycle costs based on redundancy analysis methods and electricity load change trend characteristics to obtain the equipment capacity redundancy rate.

[0011] Furthermore, the step of optimizing the cable branch box procurement list based on the equipment capacity redundancy rate and a preset cable branch box type library to obtain an optimized cost procurement list includes: selecting cable branch box type parameters from the cable branch box type library based on the equipment capacity redundancy rate and preset equipment selection conditions; performing error analysis on the predicted full-cycle cost and the calculated full-cycle cost to obtain the overall average error; and optimizing the cable branch box procurement list based on the cable branch box type parameters and the overall average error to obtain an optimized cost procurement list.

[0012] Furthermore, the cost optimization device based on cable branch boxes includes: a data acquisition module for acquiring equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters; a cost analysis module for performing cost analysis on the equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain predicted full-cycle costs and calculated full-cycle costs; a redundancy analysis module for performing redundancy analysis on the predicted full-cycle costs and calculated full-cycle costs according to a preset redundancy analysis method to obtain the equipment capacity redundancy rate; and a cost optimization module for acquiring a cable branch box procurement list and optimizing the cable branch box procurement list according to the equipment capacity redundancy rate and a preset cable branch box type library to obtain an optimized cost procurement list.

[0013] Furthermore, the cost optimization device based on cable branch boxes includes: a memory and at least one processor, wherein the memory stores instructions; at least one processor invokes the instructions in the memory to cause the cost optimization device based on cable branch boxes to perform the various steps of the cost optimization method based on cable branch boxes as described above.

[0014] Furthermore, a computer-readable storage medium stores instructions that, when executed by a processor, implement the steps of the cost optimization method based on cable branch boxes as described above.

[0015] In the technical solution of this invention, five types of parameters are collected: equipment procurement, transportation and installation, operation and maintenance, power loss, and fault repair. Based on these parameters, dynamic prediction and calculation of the entire life cycle cost are completed, forming a cost comparison basis. By comparing these two cost data through a preset redundancy analysis method, the equipment capacity redundancy rate is quantified, which can effectively identify the problem of idle equipment capacity. Finally, the original cable branch box procurement table is optimized by combining the equipment capacity redundancy rate with the cable branch box type library, and the appropriate specification box is matched as needed. This can eliminate the ineffective investment caused by equipment capacity redundancy, simplify procurement configuration, reduce the redundancy cost throughout the entire life cycle, improve the rationality, standardization, and economy of power distribution equipment procurement planning, and achieve refined management of power distribution material costs. Attached Figure Description

[0016] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0017] Figure 1 The first flowchart is provided for the cost optimization method based on cable branch boxes according to an embodiment of the present invention;

[0018] Figure 2 This is a second flowchart of a cost optimization method based on cable branch boxes provided in an embodiment of the present invention;

[0019] Figure 3 This is a third flowchart of a cost optimization method based on cable branch boxes provided in an embodiment of the present invention;

[0020] Figure 4 This is a fourth flowchart of a cost optimization method based on cable branch boxes provided in an embodiment of the present invention;

[0021] Figure 5 A fifth flowchart of a cost optimization method based on cable branch boxes provided in an embodiment of the present invention;

[0022] Figure 6 A sixth flowchart of a cost optimization method based on cable branch boxes provided in an embodiment of the present invention;

[0023] Figure 7 A seventh flowchart of a cost optimization method based on cable branch boxes provided in an embodiment of the present invention;

[0024] Figure 8 A schematic diagram of the structure of the cost optimization device based on cable branch boxes provided in an embodiment of the present invention;

[0025] Figure 9 A schematic diagram of the structure of a cost optimization device based on a cable branch box provided in an embodiment of the present invention. Detailed Implementation

[0026] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0028] 101. Obtain equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters;

[0029] In this embodiment, a unified mechanism for acquiring five-dimensional parameters of full-cycle cost is established to collect quantitative indicators for the entire chain of cable branch box procurement, construction, operation and maintenance, energy consumption, and fault compensation, providing input materials for cost calculation. The overall parameter acquisition logic is as follows: synchronously connecting with the procurement ledger, construction archive records, operation and maintenance inspection platform, electricity metering system, and fault compensation database, automatically pulling standardized structured parameters, and supporting manual supplementation of missing offline data; equipment procurement parameters include the unit price of a single box, configuration quantity, and modular plug-in cost; transportation and installation parameters cover logistics loading and unloading fees, labor construction fees, and auxiliary material costs for hole pre-embedding modification; full-cycle operation and maintenance parameters collect annual inspection, replacement of vulnerable parts, and fixed expenditures for box maintenance; power loss parameters store annual power loss, regional electricity price, and seasonal loss correction coefficient; fault repair parameters record the annual average fault rate, single emergency repair cost, and power outage compensation amount; after completing parameter collection, abnormal extreme values ​​are uniformly removed and missing fields are filled to form a standardized parameter dataset, providing unified and complete underlying parameter support for subsequent two-layer cost analysis, redundancy quantification, and procurement table optimization.

[0030] 102. Conduct cost analysis on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain predicted full-cycle costs and calculate full-cycle costs;

[0031] In this embodiment, a cost analysis mechanism is set up to complete the prediction and calculation of full-cycle costs based on the same set of five types of parameters. The cost analysis logic is divided into two-level processing flow: The first level transforms the procurement demand parameters, supplier datasets, and equipment lifecycle datasets into three-dimensional time-series features and inputs them into the PSO-XGBoost regression model to optimize the model's parameters. The model integrates dynamic variables such as raw material price cycles, seasonal load fluctuations, and equipment aging over the years to deduce and generate predicted full-cycle costs that fit the actual working conditions. The second level substitutes the five types of parameters into the preset cost item calculation formula, combines them with the time value discount factor of fixed funds, distinguishes between one-time fixed expenditures and annual continuous costs, and discounts them to present value, summarizing them to obtain a unified calculation of full-cycle costs. The two types of costs are output in pairs to form dual cost data that can be compared horizontally, providing a dual reference for subsequent redundant quantitative analysis.

[0032] 103. Perform redundancy analysis on the predicted full-cycle cost and the calculated full-cycle cost according to the preset redundancy analysis method to obtain the equipment capacity redundancy rate;

[0033] In this embodiment, the preset hierarchical redundancy analysis method is based on the dual cost difference and combines dynamic load trend characteristics to complete the redundancy attribute correction, quantify the idle proportion of the box hardware configuration and generate graded equipment capacity redundancy rates. The complete redundancy judgment logic is divided into two-layer verification process: the first layer compares the predicted full-cycle cost with the calculated full-cycle cost to calculate the cost redundancy difference, and distinguishes between temporary surplus and long-term permanent invalid redundancy based on the load change trend characteristics output by the improved LSTM algorithm according to the floor electricity demand parameters, and dynamically adjusts the preset redundancy correction coefficient; the second layer combines the rated total current carrying capacity of the cable branch box and the long-term steady-state predicted load of the floor to calculate the idle capacity proportion of the box circuit, copper busbar and heat dissipation hardware, converts it into the equipment capacity redundancy rate in the range of 0-100%, and completes the level labeling according to R≤5% no redundancy, 5%<R≤15% light redundancy, 15%<R≤30% moderate redundancy, and R>30% heavy redundancy; the equipment capacity redundancy rate intuitively quantifies the idle scale of the box configuration that exceeds the actual electricity demand, and serves as the core judgment indicator for the graded optimization of the procurement table, realizing refined redundancy quantification combined with long-term electricity dynamics.

[0034] 104. Obtain the cable branch box procurement list, and optimize the cost of the cable branch box procurement list based on the equipment capacity redundancy rate and the preset cable branch box type library to obtain an optimized cost procurement list.

[0035] In this embodiment, the tiered equipment capacity redundancy rate is used as the screening criterion. The original procurement table is iteratively optimized using a standardized enclosure type library, resulting in a low-redundancy, low-cost optimized procurement table. The overall procurement table optimization execution logic is as follows: First, the original cable branch box procurement table is imported, and information on the enclosure model, procurement quantity, circuit specifications, and estimated total cost for each corridor is extracted. Then, a preset cable branch box type library is retrieved, which stores the unit price, maintenance cost, mold amortization, and full-cycle expenditure benchmark for enclosures with different circuits and current-carrying specifications. Based on the equipment capacity redundancy rate corresponding to each corridor, the table is tiered: samples with no redundancy retain their original enclosure specifications; samples with mild, moderate, and severe redundancy have their circuit configurations progressively adjusted to match smaller enclosure specifications suitable for low loads within the type library. Simultaneously, the full-cycle procurement and maintenance estimated costs for each corridor are updated based on the enclosure type library, and the total expenditure for the single project is recalculated. After all item specifications are replaced and costs are updated, an optimized cost procurement table is generated, removing capacity redundancy and reducing ineffective expenditures throughout the entire cycle. This table can be directly used for project bidding, vendor scheduling, and budget control.

[0036] In this embodiment, five types of parameters are collected: equipment procurement, transportation and installation, operation and maintenance, power loss, and fault repair. Based on these parameters, dynamic prediction and calculation of the entire life cycle cost are completed, forming a cost comparison basis. By comparing these two cost data through a preset redundancy analysis method, the equipment capacity redundancy rate is quantified, which can effectively identify the problem of idle equipment capacity. Finally, the original cable branch box procurement table is optimized by combining the equipment capacity redundancy rate with the cable branch box type library. The appropriate specifications of the boxes are matched as needed, which can eliminate the ineffective investment caused by equipment capacity redundancy, simplify procurement configuration, reduce the redundancy cost throughout the entire life cycle, improve the rationality, standardization, and economy of power distribution equipment procurement planning, and achieve refined management of power distribution material costs.

[0037] Please see Figure 2 The second embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0038] 201. Collect procurement demand parameters, supplier datasets, equipment lifecycle datasets, and historical equipment cost data;

[0039] In this embodiment, multi-dimensional basic data required for cost modeling of cable branch boxes is collected to achieve data fusion and standardization across multiple business scenarios. The procurement demand parameters record corridor load configuration, box specifications, and procurement timeline demand information. The supplier dataset includes supply chain data such as raw material quotations, manufacturer capacity, and mold amortization. The equipment lifecycle dataset covers maintenance, loss, and fault ledger information. Historical equipment cost data retains real full-cycle cost samples from previous years. By performing noise reduction, deduplication, null value removal, and time-series alignment on the raw data, a structured and standardized unified dataset is formed. This step provides comprehensive, reliable, and unbiased raw data support for subsequent feature aggregation processing, model pre-training, and full-cycle cost calculation, ensuring the data integrity and authenticity of subsequent cost modeling and cost calculation.

[0040] 202. Based on the procurement demand parameters, supplier dataset, equipment lifecycle dataset, historical equipment cost data, and preset window aggregation method, the preset PSO-XGBoost regression model is pre-trained to obtain the cost prediction model.

[0041] In this embodiment, based on a multi-source regularized dataset and a time-series window aggregation method, feature engineering and supervised pre-training of the PSO-XGBoost regression model are completed, constructing a dedicated cost prediction model adapted to power distribution equipment scenarios. The overall model training logic is as follows: based on procurement demand parameters, supplier datasets, and equipment lifecycle datasets, lagged features and market information features are mined. Combined with a preset sliding time window and window aggregation method, time-series features are statistically aggregated to generate sliding statistical features that can reflect the medium- and long-term cost fluctuation patterns, constructing a complete three-dimensional feature system. Using the aggregated multi-dimensional features as model input and historical equipment cost data as supervision labels, the PSO particle swarm optimization algorithm is used to globally optimize and iterate the hyperparameters of the XGBoost model to obtain the optimal hyperparameter combination and complete the model iterative training. After the model prediction error converges to a preset threshold, the model weights and feature mapping relationship are solidified to obtain the cost prediction model, which can adapt to market fluctuations, load time-series changes, and equipment aging patterns, providing a reliable model carrier for dynamic full-cycle intelligent cost prediction.

[0042] 203. Based on the cost prediction model, predictive analysis is performed on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain the predicted full-cycle cost;

[0043] In this embodiment, the pre-trained cost prediction model serves as a dynamic cost extrapolation vehicle, outputting a realistic prediction of the entire lifecycle cost. The overall prediction and judgment logic involves standardizing and encoding equipment procurement parameters, transportation and installation parameters, lifecycle operation and maintenance parameters, power loss parameters, and fault repair parameters, transforming them into feature inputs that the model can recognize. Utilizing the time-series feature patterns and market change mapping relationships built into the cost prediction model, and comprehensively considering dynamic influencing factors such as raw material price fluctuations, seasonal load disturbances, equipment aging over the years, and changes in fault probability, the model extrapolates and calculates the costs of the cable branch box throughout its historical lifecycle, including purchase, installation, operation and maintenance, power loss, and fault compensation. Finally, it outputs a dynamic prediction of the entire lifecycle cost, which accurately reflects the dynamic expenditure level of the equipment throughout its entire lifecycle, providing dynamic cost benchmark data for subsequent redundancy error analysis.

[0044] 204. Calculate the equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters according to the preset cost item calculation formula and the preset time value of money discount factor to obtain the full-cycle cost.

[0045] In this embodiment, a preset cost item calculation formula, combined with a fixed time value of money discount factor, generates a benchmark for calculating the full-cycle cost required for redundancy analysis. This benchmark is then used to effectively compare the full-cycle cost with the predicted cost, providing a unified judgment standard for equipment capacity redundancy analysis and redundancy rate quantification, thereby improving the accuracy of the full-cycle cost assessment of power distribution equipment. The complete calculation execution logic, the full set of calculation formulas, and parameter definitions are as follows:

[0046] The expressions for the sub-formulas are as follows:

[0047] One-time purchase cost of equipment: ,

[0048] Total one-time cost of transportation and installation: ,

[0049] Maintenance costs for year t: ,

[0050] Electricity loss cost in year t: ,

[0051] Fault repair compensation costs in year t: ,

[0052] In the formula, To calculate the total cost over the entire lifecycle, the historical lifecycle discounted total cost of the cable branch box is calculated. This refers to the one-time purchase cost of the equipment. This refers to the total one-time cost of transporting, installing, or modifying the equipment. The standardized operation and maintenance cost for year t; The economic cost of the enclosure under no-load and heat loss in year t; The total cost of fault repair and power outage compensation in year t; T is the standard full life cycle of the equipment, fixed at 15 years. The preset time value of money discount factor is used to discount future annual costs to their present value in the current period; N represents the standard unit price for a single cable distribution box (from equipment procurement parameters); N represents the number of cable distribution boxes configured in a single building corridor (from equipment procurement parameters). The cost of transporting and unloading a single piece of equipment (derived from transport and installation parameters); This refers to the on-site labor costs for installation (derived from transportation and installation parameters). Costs of auxiliary materials for drilling, pre-embedding, and sealing modifications (from transportation and installation parameters); The annual inspection labor cost (derived from full-cycle operation and maintenance parameters); Cost of replacing vulnerable parts such as plugs and seals annually (from full lifecycle maintenance parameters); Fixed annual maintenance expenses for enclosure corrosion protection and fastening (derived from full-cycle operation and maintenance parameters); The total equivalent annual power loss of the enclosure (power loss parameter). For a unified regional electricity price (based on power loss parameters); This is the seasonal loss correction factor (power loss parameter). The annual average failure rate of the corresponding model enclosure (from failure repair parameters); Cost of replacing and repairing a single faulty component (from fault repair parameters); Economic compensation to residents and additional rework costs caused by power outages (from fault repair parameters);

[0053] In this embodiment, a full-cycle cost accounting system combining multi-source data normalization, temporal feature optimization, model prediction, and calculation is constructed. By collecting and standardizing multi-dimensional data on procurement, suppliers, equipment lifecycle, and historical costs, the integrity and reliability of modeling and calculation data are ensured. Multi-dimensional temporal features are mined using the window aggregation method, and supervised training is completed using the PSO-XGBoost model with particle swarm optimization, effectively adapting to market price fluctuations, load temporal changes, and equipment aging patterns, thereby improving the accuracy of dynamic cost prediction. At the same time, cost accounting is carried out by using itemized cost calculation formulas combined with the time value of money discount factor. By comparing the dynamic prediction of full-cycle costs with the calculated full-cycle costs, a unified cost assessment benchmark is formed, providing a basis for quantifying equipment capacity redundancy, error analysis, and optimizing the power distribution equipment procurement list, effectively improving the refinement of the full-cycle cost calculation of power distribution materials.

[0054] Please see Figure 3 The third embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0055] 301. Based on the preset feature analysis method, perform feature analysis on the procurement demand parameters, supplier datasets and equipment life cycle datasets to obtain lag features and market information features;

[0056] In this embodiment, the preset feature analysis method sets up multi-dimensional feature extraction and filtering logic to complete feature decomposition, association matching, and redundancy removal for three types of original datasets: procurement demand, suppliers, and equipment lifecycle. It outputs lag features and market information features for model training. The overall feature extraction judgment logic is as follows: iterates through the procurement demand parameters (number of households in buildings, load planning, procurement cycle time series records), supplier datasets (raw material quotations, production capacity, mold amortization time series data), and equipment lifecycle maintenance, fault, and loss history records, classifying features according to time series delay attributes. The system identifies several characteristics: load, raw material prices, and historical failure indicators with time-lag transmission effects as lagging features, as past values ​​of these indicators will have a long-term impact on current tank purchase and maintenance costs; real-time indicators such as daily raw material market prices, manufacturer price adjustments, new building plans, and renovation policies are identified as market information features; multicollinearity and irrelevant invalid fields are simultaneously removed, feature normalization encoding is completed, and two types of standardized features are separated and output to provide basic feature materials for subsequent sliding window aggregation statistics and model training, thus fully mining the time-series influencing factors of the entire procurement, supply, and maintenance chain.

[0057] 302. Based on the preset sliding time window and window aggregation method, aggregate and statistically analyze the lag characteristics and market information characteristics to obtain sliding statistical characteristics;

[0058] In this embodiment, multiple preset sliding time windows, combined with window aggregation operation rules, are used to perform periodic quantification on two types of time-series features: lagging features and market information features, generating sliding statistical features that reflect medium- and long-term fluctuation patterns. The overall aggregation execution logic is as follows: three fixed sliding time windows are set: 3 months for short-term, 6 months for medium-term, and 12 months for long-term. The sliding step size is uniformly set to a single month. The window is shifted forward month by month, discarding the oldest time-series data and incorporating the latest indicators. For all feature values ​​within each window, five types of aggregation indicators are calculated simultaneously: mean, peak, minimum, standard deviation, and cumulative total over the period. Loops, apartment types, and operation and maintenance environments are grouped and aggregated independently, without cross-scenario mixed calculations. All window aggregation indicators are normalized and uniformly generated to generate sliding statistical features that depict the seasonal cyclical changes in raw material prices, corridor load, and fault losses, making up for the randomness defects of single-point-of-time features and improving the model's three-dimensional input feature system.

[0059] 303. The PSO-XGBoost regression model is pre-trained based on historical equipment cost data, sliding statistical characteristics, lag characteristics, and market information characteristics to obtain a cost prediction model;

[0060] In this embodiment, historical equipment cost data is used as a supervisory label to complete the pre-training of the PSO-XGBoost regression model, generating a dedicated cost prediction model for cable branch boxes. The complete training logic consists of a two-level process: The first level inputs all three-dimensional features—sliding statistical features, lag features, and market information features—into the PSO particle swarm optimization module. Using cost prediction error as the fitness index, it globally traverses the XGBoost hyperparameter combinations, iteratively updating the local optimum of particles and the global optimum of the population hyperparameter combinations, and outputs the optimized regression model. The second level hierarchically divides the historical equipment procurement, installation, operation and maintenance, and failure full-cycle cost dataset into training and validation sets. The three-dimensional features are used as input, and the real LCC cost is used as a label to feed into the optimized regression model for iterative training. The prediction error of the validation set is continuously monitored. After the error converges to a preset error threshold, the model weights and feature mapping relationship are solidified to obtain the final cost prediction model. This model can dynamically output the full-cycle predicted cost of the corridor box, providing a quantitative calculation tool for subsequent dual cost redundancy analysis and procurement table optimization.

[0061] In this embodiment, a multi-source data feature extraction, time-series window aggregation, and particle swarm optimization model training process is established. Feature analysis is used to distinguish between lagging features and market news features from three types of data: procurement demand, suppliers, and equipment lifecycle data. This fully uncovers the impact of time-series lag and real-time market changes on the cost of cable distribution boxes. Time-series aggregation statistics are completed using multi-period sliding time windows to generate sliding statistical features, smoothing out short-term random fluctuations and capturing the medium- and long-term cyclical patterns of load, raw materials, and fault losses, thus constructing a complete three-dimensional feature input system. Based on the three-dimensional features and historical equipment cost data, global hyperparameter optimization and supervised pre-training are performed on the PSO-XGBoost model, automatically converging the optimal model weights and improving the accuracy of full-cycle cost prediction. The model's dynamically predicted cost output can support subsequent capacity redundancy quantification and procurement list optimization, effectively reducing redundant expenditures on cable branch box purchase, maintenance, and renovation, and improving the precision of power distribution material cost calculation.

[0062] Please see Figure 4 The fourth embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0063] 401. Based on the lag characteristics, sliding statistical characteristics, and market information characteristics, the parameters of the PSO-XGBoost regression model are tuned to obtain the optimized regression model;

[0064] In this embodiment, the preset PSO particle swarm optimization mechanism serves as a global hyperparameter optimization constraint. It leverages three types of corridor-specific features to complete the global parameter iteration and tuning of the XGBoost regression model, outputting a tuned regression model adapted to the power distribution cost calculation scenario. The overall parameter tuning logic is as follows: The model's complete input feature set is constructed using lag features, sliding statistics features, and market information features; all hyperparameters, including XGBoost tree depth, learning rate, sampling ratio, and regularization coefficient, are used as the particle population dimension, with each particle corresponding to a complete hyperparameter combination; the preset reciprocal of the cost prediction MAPE error is used as the particle... The fitness index iteratively updates the local optimal and global optimal hyperparameter combinations of particles, while setting parameter boundary constraints to avoid extreme parameters with no business significance. When the iteration reaches the maximum number of generations or the accuracy has not improved for several consecutive rounds, the optimization terminates, the global optimal hyperparameters are extracted, the XGBoost base network is reconfigured, and a convergent optimal tuned regression model is generated. This tuned regression model eliminates the accuracy deviation caused by manual parameter tuning, and fully adapts to three types of influencing factors: load time-series fluctuations, raw material price cycles, and market policy disturbances. It provides a high-precision base network carrier for subsequent training of standardized cost prediction models based on historical cost data.

[0065] 402. Pre-train the optimized regression model based on historical equipment cost data to obtain a cost prediction model;

[0066] In this embodiment, the optimized regression model serves as the base network. Supervised pre-training is conducted using historical equipment cost data from the entire sample as supervised labels to generate a dedicated cost prediction model capable of real-time calculation of the entire lifecycle expenditure of the enclosure. The complete training execution logic is as follows: the historical equipment procurement, installation, operation and maintenance, and failure loss full-cycle cost data, after stratification and cleaning, are divided into training and validation sets. The regression model is iteratively trained using three-dimensional features—lag, sliding, and market information—as input, and the corresponding historical real LCC cost as the label input. During the training process, the prediction error of the validation set is continuously monitored. Training stops when the error converges to a preset error threshold, and the model weights and feature mapping relationships are solidified to obtain the final cost prediction model. This model can predict and analyze the full-cycle predicted cost, providing a dynamic cost calculation tool for subsequent cost comparison, redundancy analysis, and procurement optimization.

[0067] In this embodiment, multi-dimensional features combined with the PSO particle swarm optimization algorithm are used to automatically optimize the parameters of the XGBoost model. By leveraging hysteresis features, sliding statistics features, and market information features, the model adapts to the dynamic changes in power distribution costs, improving the adaptability and accuracy of the model parameter configuration. Furthermore, supervised pre-training is conducted using massive amounts of standardized and cleaned historical equipment lifecycle cost data. Through hierarchical iterative training of the dataset, error convergence is achieved, and model weights are solidified to construct a high-precision dedicated cost prediction model. This model can capture the impact of load time-series fluctuations, raw material price changes, and market policy disturbances on cable branch box costs, enabling intelligent prediction of dynamic costs throughout the entire lifecycle. The overall model training and parameter configuration system is optimized, reducing cost prediction errors and providing accurate and reliable data support for subsequent dual-cost comparisons, equipment capacity redundancy analysis, and procurement table optimization, effectively improving the accuracy of the full lifecycle cost calculation for power distribution equipment.

[0068] Please see Figure 5 The fifth embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0069] 501. Determine whether the calculated total lifecycle cost is greater than the preset lifecycle cost threshold;

[0070] In this embodiment, the preset cycle cost threshold is the historical full life cycle benchmark cost limit of a single corridor cable branch box, which serves as a pre-judgment switch to trigger redundancy analysis and distinguish whether load and redundancy quantification processes need to be carried out. The overall judgment logic is as follows: read the calculated full cycle cost of the corresponding box in a single corridor and compare it with the pre-stored standard cycle cost threshold for the same circuit and the same apartment type. If the calculated full cycle cost is less than or equal to the cycle cost threshold, it means that the overall expenditure on box purchase and operation and maintenance is within the standard reasonable range and there is no obvious capacity redundancy. The subsequent power consumption parameter collection and redundancy analysis process is skipped directly. If the calculated full cycle cost is greater than the cycle threshold, it means that the corridor box is likely to have redundant losses caused by excessive circuit and current carrying capacity configuration. The downstream data collection command is automatically triggered to retrieve the power consumption parameters of the complete floor of the corridor, providing the underlying load input material for the subsequent redundancy quantification analysis, realizing the on-demand activation of redundancy calculation and reducing the meaningless computing power consumption of the system.

[0071] 502. If the calculated total lifecycle cost is greater than the lifecycle cost threshold, then obtain the floor's electricity demand parameters;

[0072] In this embodiment, the threshold for calculating the full-cycle cost is used as the trigger condition for collecting electricity consumption parameters. Only when the overall cost of the enclosure is determined to exceed the standard is the corresponding building's full-dimensional electricity consumption time-series data automatically retrieved. The collection scope covers all floor electricity demand parameters, including monthly electricity consumption of the enclosure, rated load of each household, annual occupancy changes, seasonal temperature and humidity, and peak electricity consumption time series. The data is simultaneously bound to the enclosure number and enclosure circuit label to complete data standardization and cleaning, removing load null values ​​and extremely distorted time series records, and uniformly organizing them into a structured time series dataset. This set of electricity demand parameters is the underlying data source for redundancy analysis and load trend prediction, providing dynamic load basis for subsequent hierarchical redundancy calculations, avoiding invalid data collection in scenarios with no cost anomalies, and improving the computational efficiency of the entire cost optimization process.

[0073] 503. Based on the redundancy analysis method and the floor power demand parameters, perform redundancy analysis on the calculated full-cycle cost and the predicted full-cycle cost to obtain the equipment capacity redundancy rate.

[0074] In this embodiment, redundancy analysis is conducted by combining floor power demand parameters. Redundancy quantification is completed by comparing two types of full-cycle costs: calculated full-cycle cost and predicted full-cycle cost. Based on real corridor time-series power consumption data, seasonal temporary load surplus and long-term equipment capacity idleness are distinguished. Redundancy calculation coefficients are dynamically corrected to calculate the graded equipment capacity redundancy rate. The idle loss of box circuits and copper busbar configurations is intuitively quantified, providing reliable graded indicators for cable branch box selection and procurement table optimization. This reduces the redundant expenditures on procurement, operation and maintenance, and renovation caused by large-capacity boxes, and improves the accuracy of full-cycle cost calculation of power distribution materials and the implementation effect of procurement optimization.

[0075] In this embodiment, a cost analysis system based on threshold screening and dynamic redundancy quantification is constructed. This effectively compensates for the shortcomings of blind redundancy judgment and wasted computing power in the cost control of cable branch boxes. Using a preset periodic cost threshold as a pre-judgment condition, abnormal box samples with excessive full-cycle costs are screened out. Subsequent analysis processes are only initiated for targets with redundant losses, reducing invalid system calculations and improving overall computing efficiency. At the same time, time-series parameters of floor electricity demand are collected in a targeted manner to provide real and reliable data support for redundancy analysis. By completing cost redundancy analysis based on real electricity consumption data, seasonal temporary load surplus and long-term equipment capacity idleness can be effectively distinguished. The calculation coefficient is dynamically corrected, the redundancy loss of box hardware configuration is quantified, and the equipment capacity redundancy rate is generated. This provides a quantitative basis for box model selection and procurement table optimization, reducing redundant costs in equipment procurement, operation and maintenance, and renovation from the source, and improving the accuracy of full-cycle cost calculation of power distribution equipment.

[0076] Please see Figure 6 The sixth embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0077] 601. Based on a preset improved LSTM time series prediction algorithm and a preset seasonal fluctuation coefficient, the electricity demand parameters of each floor are predicted to obtain the trend characteristics of electricity load changes.

[0078] In this embodiment, the preset improved LSTM time series prediction algorithm uses the original time series parameters of electricity demand for all floors in the building as input, integrates seasonal fluctuation coefficients to complete the time series load extrapolation, and outputs long-term electricity load change trend characteristics that can be used for redundancy determination. The overall prediction execution logic is as follows: first, import the original floor electricity demand parameters such as historical monthly floor electricity consumption, occupancy rate, temperature and humidity, input the time series samples into the improved LSTM basic network to extract the basic time series patterns of the load, and then superimpose the preset seasonal fluctuation coefficients for spring, summer, autumn and winter to make weighted corrections for peak and valley loads, offsetting the prediction deviations caused by seasonal electricity disturbances of air conditioning and heating; the improved LSTM... Internally, a load forgetting gate and a long-term trend memory gate are added to distinguish between short-term monthly fluctuations and medium-to-long-term load growth patterns over 3-12 months. Simultaneously, five trend indicators are output: annual year-on-year growth slope of floor load, average summer peak load, spring and autumn stable load benchmark, seasonal fluctuation amplitude, and the occupancy load growth rate for the next three years. These five indicators combine to form the electricity load change trend characteristics. This set of electricity load change trend characteristics serves as a pre-input for redundancy analysis, distinguishing between short-term temporary load fluctuations and long-term steady-state load scenarios. This avoids misjudgments caused by relying solely on static rated load for redundancy determination, providing a quantitative basis for dynamic load changes in subsequent redundancy analysis.

[0079] 602. Based on the redundancy analysis method and the characteristics of electricity load change trends, perform redundancy analysis on the calculation of full-cycle cost and the prediction of full-cycle cost to obtain the equipment capacity redundancy rate;

[0080] In this embodiment, the preset hierarchical redundancy analysis method adjusts the preset redundancy correction coefficient based on the dynamic electricity load trend characteristics, and simultaneously compares the calculated full-cycle cost and the dynamically predicted full-cycle cost to quantify the cost redundancy caused by the idle capacity of the enclosure in a hierarchical manner, and finally outputs the quantified equipment capacity redundancy rate; the complete redundancy judgment logic is divided into two layers of verification process: the first layer distinguishes redundancy attributes based on the electricity load change trend characteristics. If the annual growth slope of the electricity load change trend characteristics is positive and the occupancy continues to increase, it is judged as short-term phased redundancy and the redundancy correction coefficient is lowered; if the load is stable in the long term or shrinks year by year, it is judged as permanent invalid redundancy and the redundancy correction coefficient is raised; the second layer uses the trend correction coefficient to compare the calculated cost with the predicted full-cycle cost. The cost difference is converted to obtain the cost redundancy. Then, combined with the rated total current carrying capacity of the cable branch box and the long-term steady-state predicted load of the floor, the idle capacity ratio of the box is calculated and converted into the equipment capacity redundancy rate in the range of 0-100%. The system simultaneously completes the level labeling according to R≤5% no redundancy, 5%<R≤15% light redundancy, 15%<R≤30% moderate redundancy, and R>30% heavy redundancy. The equipment capacity redundancy rate fully quantifies the degree of idleness of the box circuit, copper busbar, and heat dissipation configuration exceeding the actual dynamic load. It serves as a graded judgment indicator for subsequent cable branch box model selection and procurement list optimization, realizing refined redundancy quantification combined with long-term power consumption dynamics, and solving the defect of static cost accounting that cannot identify seasonal and phased capacity idleness.

[0081] In this embodiment, an improved LSTM time-series prediction algorithm combined with seasonal fluctuation coefficients is used to predict floor electricity consumption, distinguishing between short-term fluctuations and medium- to long-term load growth patterns, and outputting the trend characteristics of electricity load changes. This strengthens the data foundation for redundancy analysis from the perspective of dynamic electricity consumption. Based on a preset hierarchical redundancy analysis method, the preset redundancy correction coefficient is adjusted according to the trend characteristics of dynamic electricity load. The cost redundancy is calculated by combining the cost difference between the two types of full-cycle costs, and then the capacity redundancy rate of graded equipment is converted to quantify the idle degree of box circuits and current-carrying configurations. The entire two-layer analysis system solves the pain point of traditional static calculations being prone to misjudging seasonal temporary capacity surplus. The redundancy grading results objectively conform to the long-term change patterns of community occupancy and electricity consumption, providing a quantitative grading basis for cable branch box selection and procurement list optimization. This avoids the multiple ineffective costs of molds, operation and maintenance, and renovation caused by blindly purchasing large-capacity boxes, improves the refinement and accuracy of the full-cycle cost calculation of power distribution materials, and strengthens the procurement optimization plan.

[0082] Please see Figure 7 The seventh embodiment of the cost optimization method based on cable branch boxes in this invention includes:

[0083] 701. Select cable branch box type parameters from the cable branch box type library based on the equipment capacity redundancy rate and preset equipment selection conditions;

[0084] In this embodiment, the preset equipment selection conditions are hierarchical matching and screening constraints. Combined with the equipment capacity redundancy rate, the internal specification matching and compatibility verification logic of the cable branch box type library is completed. The overall selection judgment logic is as follows: the equipment capacity redundancy rate corresponding to the current box in a single corridor / area matches the corresponding load range selection standard, and the rated current carrying capacity, outgoing circuits, and hardware configuration of the candidate box simultaneously meet the preset multi-dimensional selection constraints before the complete type parameters of the model can be extracted. The system first reads the equipment capacity redundancy rate R output from the previous redundancy analysis and divides it into four levels of selection criteria according to redundancy level: R≤5% (no redundancy range) uses the existing same specification selection conditions; 5%<R≤15% (mild redundancy) lowers the circuit configuration selection standard by one level; 15%<R≤30% (moderate redundancy) forcibly lowers the load compatibility selection standard by two levels; R>30% (severe redundancy) replaces the minimum compatible box model, and the preset equipment selection conditions are retrieved simultaneously. The equipment selection conditions include... Five key matching dimensions: long-term steady-state load limit of corridors, annual load growth slope, building installation size limit, environmental corrosion resistance level, and full-cycle cost benchmark threshold; using redundancy level as the first-level filtering index, the system traverses all model files in the pre-stored cable branch box type library, comparing each model's rated capacity, outgoing circuits (6 / 9 / 12 circuits), modular base configuration, and single-unit LCC benchmark cost to see if they simultaneously match the selection constraints corresponding to the current redundancy level; eliminating box models with excessive capacity configuration, incompatible sizes, or excessive long-term operation and maintenance costs, retaining the best-fit single specification, and fully reading and caching all cable branch box type parameters corresponding to that model, including circuit specifications, base structure, detachable plug configuration, single-unit purchase price, transportation and installation costs, average annual operation and maintenance losses, and standard service life; completing the model matching and filtering, providing standardized box specifications and cost benchmark parameters to support subsequent cost optimization of the procurement table;

[0085] 702. Conduct error analysis between the predicted full-cycle cost and the calculated full-cycle cost to obtain the overall average error;

[0086] In this embodiment, the preset error analysis calculation logic is a hierarchical sample deviation statistical rule. Using a single corridor and a single area of ​​enclosures as independent sample units, it compares two sets of values: the predicted full-cycle cost and the calculated full-cycle cost for enclosures of the same specification. After calculating the single-sample deviation hierarchically, it summarizes the overall comprehensive index and outputs the overall average error as the procurement optimization adjustment coefficient. The system first distinguishes between two sets of cost definitions: the predicted full-cycle cost is the dynamic full-lifecycle LCC cost output by PSO-XGBoost integrating lag, sliding, and market information features; the calculated full-cycle cost is the cost calculated based on the cost item calculation formula and the time value of money discount factor. For each corridor sample, it first calculates the single-sample absolute deviation and the relative... For deviations, outlier samples caused by extreme loads or temporary raw material price spikes are filtered out. Then, the average relative error of each group is statistically analyzed according to the 6-loop N1, 9-loop N2, and 12-loop N3 enclosures. Finally, the deviation values ​​of all samples in the entire domain are weighted to obtain the globally unified overall average MAP error (overall average error). The overall average error is used to quantify the degree of deviation in the cost model. The higher the error value, the more serious the distortion in the matching between the current load, enclosure configuration and cost. This indicator serves as a floating adjustment factor for procurement optimization, dynamically adjusting the upper and lower fluctuation ranges of the procurement batch size for various enclosures in the subsequent procurement list optimization process, constraining the procurement ratio of large-capacity, high-redundancy enclosures, and providing a quantitative basis for cost reduction and optimization of the procurement list.

[0087] 703. Optimize the cost of the cable branch box procurement list based on the cable branch box type parameters and the overall average error to obtain an optimized cost procurement list;

[0088] In this embodiment, this step is the final optimization output stage of the procurement list, simultaneously incorporating two adjustment parameters: the optimal enclosure type parameter obtained through equipment capacity redundancy rate screening and the overall average error; setting a two-layer procurement optimization constraint logic, the first layer replaces the high-redundancy large-specification enclosures in the original procurement list based on the enclosure type parameter, and the second layer dynamically adjusts the floating margin of safety stock for various enclosures based on the overall average error, correcting the original procurement list to achieve comprehensive cost reduction; the first layer optimization logic: reads the original procurement enclosure model and procurement quantity for each building and corridor in the original cable branch box procurement list; calls the optimal matching enclosure type parameter, compares the original model capacity redundancy level, and batch replaces the 12-circuit and 9-circuit large-capacity procurement items corresponding to medium and heavy redundancy with low-redundancy 6 / 9-circuit standard models that match the corridor load, and simultaneously updates the single unit full-cycle procurement and maintenance benchmark cost based on the type parameter, initially reducing the long-term ineffective expenditure caused by redundant equipment; the second layer optimization logic: reads the calculation from step B The calculated overall average error is used to set floating margin matching rules: if the overall average error is ≤5% (high model accuracy), only a 2% safety buffer is retained for the purchase quantity of various types of enclosures; if 5% < overall average error ≤8% (moderate deviation), the buffer margin is increased to 5%; if the overall average error >8% (large model deviation), the buffer margin is increased by 8%. The total purchase quantity of enclosures for each loop is adjusted accordingly to balance the risk of material shortage and the loss of inventory capital backlog. During the optimization process, the three hard boundaries of the total project budget, manufacturer modular capacity, and zoned warehousing are checked simultaneously. The total purchase quantity after any adjustment must not exceed the preset constraints. After all items have been replaced with models and the purchase quantity has been corrected, the total purchase life cycle comprehensive cost of the entire region is re-summarized to generate a new form that removes capacity redundancy and takes into account the supply and demand balance, namely the optimized cost purchase form. The form is marked by building, including the optimal enclosure model, cost per LCC, batch purchase plan, and annual mold saving calculation value, which can be directly used for material bidding and manufacturer modular production scheduling.

[0089] In this embodiment, the cable distribution box specifications are screened based on equipment capacity redundancy rate. Combined with multi-dimensional equipment selection criteria such as corridor load, building dimensions, and corrosion resistance, suitable models are matched from the cable distribution box type library. Large-capacity redundant boxes are eliminated, reducing ineffective costs associated with equipment purchase, mold amortization, and long-term unloaded maintenance, thus solving the resource waste problem caused by the unified procurement of large-specification boxes. By comparing the dynamically predicted full-cycle cost with the calculated full-cycle cost, stratified error statistics are conducted, and the overall average error is output as a quantitative adjustment factor to measure the accuracy of the cost model calculation, dynamically adjusting the procurement safety stock, and balancing shortages. The system addresses the issues of expedited pricing and inventory capital tied up in inventory. By optimizing the original procurement list through the linkage of optimal container parameters and error index constraints, it replaces highly redundant procurement items in batches, adjusts procurement buffers at different levels, and simultaneously verifies budget, capacity, and warehousing constraints. The result is a standardized and optimized procurement list, reducing the overall expenditure throughout the entire lifecycle of the corridor branch boxes. The entire logic is interconnected, achieving load matching with container capacity, improving the level of refined management of material procurement, reducing hidden losses such as mismatched modifications and idle equipment, and providing manufacturers with production scheduling basis for modular production, thus balancing procurement economy and supply stability.

[0090] The cost optimization method based on cable branch boxes in the embodiments of the present invention has been described above. The cost optimization device based on cable branch boxes in the embodiments of the present invention will be described below. Please refer to [link / reference]. Figure 8 One embodiment of the cost optimization device based on cable branch boxes in this invention includes:

[0091] Data acquisition module 1 is used to acquire equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters;

[0092] Cost analysis module 2 is used to perform cost analysis on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters and fault repair parameters, so as to obtain the predicted full-cycle cost and the calculation of the full-cycle cost;

[0093] Redundancy analysis module 3 is used to perform redundancy analysis on the predicted full-cycle cost and the calculated full-cycle cost according to the preset redundancy analysis method, so as to obtain the equipment capacity redundancy rate.

[0094] The cost optimization module 4 is used to obtain the cable branch box purchase list and optimize the cost of the cable branch box purchase list according to the equipment capacity redundancy rate and the preset cable branch box type library to obtain an optimized cost purchase list.

[0095] In this embodiment, five types of parameters are collected: equipment procurement, transportation and installation, operation and maintenance, power loss, and fault repair. Based on these parameters, dynamic prediction and calculation of the entire life cycle cost are completed, forming a cost comparison basis. By comparing these two cost data through a preset redundancy analysis method, the equipment capacity redundancy rate is quantified, which can effectively identify the problem of idle equipment capacity. Finally, the original cable branch box procurement table is optimized by combining the equipment capacity redundancy rate with the cable branch box type library. The appropriate specifications of the boxes are matched as needed, which can eliminate the ineffective investment caused by equipment capacity redundancy, simplify procurement configuration, reduce the redundancy cost throughout the entire life cycle, improve the rationality, standardization, and economy of power distribution equipment procurement planning, and achieve refined management of power distribution material costs.

[0096] Figure 9 This is a schematic diagram of the structure of a cost optimization device based on a cable branch box provided in an embodiment of the present invention. The cost optimization device 900 based on the cable branch box can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 910 (e.g., one or more processors) and a memory 920, and one or more storage media 930 (e.g., one or more mass storage devices) for storing application programs 933 or data 932. The memory 920 and storage media 930 can be temporary or persistent storage. The program stored in the storage media 930 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the cost optimization device 900 based on the cable branch box. Furthermore, the processor 910 may be configured to communicate with the storage media 930 and execute the series of instruction operations in the storage media 930 on the cost optimization device 900 based on the cable branch box to implement the steps of the cost optimization method based on the cable branch box provided in the above-described method embodiments.

[0097] The cost optimization device 900 based on the cable distribution box may also include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input / output interfaces 960, and / or one or more operating systems 931, such as Windows Server, MacOSX, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 9 The illustrated cost optimization device structure based on cable branch boxes does not constitute a limitation on the cost optimization device based on cable branch boxes. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0098] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of a cost optimization method based on a cable branch box.

[0099] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0100] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0101] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A cost optimization method based on cable branch boxes, characterized in that, include: Obtain equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters; Cost analysis is conducted on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain predicted and calculated full-cycle costs. Redundancy analysis is performed on the predicted and calculated full-cycle costs based on a preset redundancy analysis method to obtain the equipment capacity redundancy rate. Obtain the cable branch box procurement list, and optimize the cost of the cable branch box procurement list based on the equipment capacity redundancy rate and the preset cable branch box type library to obtain an optimized cost procurement list.

2. The cost optimization method based on cable branch boxes as described in claim 1, characterized in that, The cost analysis of equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain predicted and calculated full-cycle costs includes: Collect procurement demand parameters, supplier datasets, equipment lifecycle datasets, and historical equipment cost data; The pre-trained PSO-XGBoost regression model is obtained by using procurement demand parameters, supplier datasets, equipment lifecycle datasets, historical equipment cost data, and a pre-defined window aggregation method. Based on the cost prediction model, predictive analysis is performed on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters to obtain the predicted full-cycle cost. Based on the preset cost item calculation formula and the preset time value of money discount factor, the equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters and fault repair parameters are calculated to obtain the full-cycle cost.

3. The cost optimization method based on cable branch boxes as described in claim 2, characterized in that, The process of pre-training a pre-defined PSO-XGBoost regression model based on procurement demand parameters, supplier datasets, equipment lifecycle datasets, historical equipment cost data, and a pre-defined window aggregation method to obtain a cost prediction model includes: Based on the preset feature analysis method, feature analysis is performed on the procurement demand parameters, supplier datasets and equipment life cycle datasets to obtain lag features and market information features; Based on the preset sliding time window and window aggregation method, the lag characteristics and market information characteristics are aggregated and statistically analyzed to obtain sliding statistical characteristics; The PSO-XGBoost regression model was pre-trained based on historical equipment cost data, sliding statistics, lag characteristics, and market information characteristics to obtain a cost prediction model.

4. The cost optimization method based on cable branch boxes as described in claim 3, characterized in that, The process of pre-training the PSO-XGBoost regression model based on historical equipment cost data, sliding statistical characteristics, lag characteristics, and market information features to obtain a cost prediction model includes: The parameters of the PSO-XGBoost regression model were tuned based on lag characteristics, sliding statistics characteristics, and market information characteristics to obtain the optimized regression model. The regression model is pre-trained based on historical equipment cost data to obtain a cost prediction model.

5. The cost optimization method based on cable branch boxes as described in claim 1, characterized in that, The step of performing redundancy analysis on the predicted and calculated full-cycle costs according to a preset redundancy analysis method to obtain the equipment capacity redundancy rate includes: Determine whether the calculated total lifecycle cost exceeds the preset lifecycle cost threshold; If the calculated total lifecycle cost is greater than the lifecycle cost threshold, then obtain the floor's electricity demand parameters; Redundancy analysis is performed on the calculated and predicted full-cycle costs based on redundancy analysis methods and floor power demand parameters to obtain the equipment capacity redundancy rate.

6. The cost optimization method based on cable branch boxes as described in claim 5, characterized in that, The redundancy analysis is performed on the calculated and predicted full-cycle costs based on redundancy analysis methods and floor power demand parameters to obtain the equipment capacity redundancy rate, including: Based on a preset improved LSTM time series prediction algorithm and a preset seasonal fluctuation coefficient, the electricity demand parameters of each floor are predicted to obtain the trend characteristics of electricity load changes. Redundancy analysis is performed on the calculation and prediction of full-cycle costs based on redundancy analysis methods and the characteristics of electricity load change trends to obtain the equipment capacity redundancy rate.

7. The cost optimization method based on cable branch boxes as described in claim 2, characterized in that, The process of optimizing the cable branch box procurement list based on equipment capacity redundancy rate and a pre-set cable branch box type library to obtain an optimized cost procurement list includes: The cable branch box type parameters are selected from the cable branch box type library based on the equipment capacity redundancy rate and preset equipment selection conditions. An error analysis is performed between the predicted full-cycle cost and the calculated full-cycle cost to obtain the overall average error; The cable branch box procurement list is optimized based on the cable branch box type parameters and the overall average error to obtain an optimized cost procurement list.

8. A cost optimization device based on a cable branch box, characterized in that, include: The data acquisition module is used to acquire equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters; The cost analysis module is used to perform cost analysis on equipment procurement parameters, transportation and installation parameters, full-cycle operation and maintenance parameters, power loss parameters, and fault repair parameters in order to obtain predicted full-cycle costs and calculate full-cycle costs. The redundancy analysis module is used to perform redundancy analysis on the predicted full-cycle cost and the calculated full-cycle cost according to the preset redundancy analysis method, so as to obtain the equipment capacity redundancy rate. The cost optimization module is used to obtain the cable branch box procurement list and optimize the cost of the cable branch box procurement list based on the equipment capacity redundancy rate and the preset cable branch box type library to obtain an optimized cost procurement list.

9. A cost optimization device based on cable branch boxes, characterized in that, The cost optimization device based on cable branch box includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the cable branch box-based cost optimization device to perform the steps of the cable branch box-based cost optimization method as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that, When the instructions are executed by the processor, they implement the steps of the cost optimization method based on cable branch boxes as described in any one of claims 1-7.

Images