Dynamic pricing optimization method and system for coal industry chain, and program product

Abstract

The invention discloses a dynamic pricing optimization method and system for a coal industry chain and a program product, and the method comprises the steps: automatically obtaining multi-source data of the coal industry chain through a data collection and fusion module, and carrying out the fusion processing of the obtained spatio-temporal data; a pricing prediction module predicts the market price trend based on the fused data, and quantifies the response relationship of the price change to the supply and demand quantity; performing prediction analysis on a supply end, a demand end and an inventory end of the industry chain based on the fused data through a supply and demand analysis module; an evaluation strategy is dynamically selected through a quantitative evaluation module according to the market environment state, and quantitative scoring is carried out on the pricing scheme based on the pricing prediction and supply and demand analysis results; and solving by adopting a multi-objective optimization algorithm based on the quantitative scoring result through an optimization solving module, and outputting an optimal pricing scheme. According to the method, collaborative optimal pricing under the condition of balancing multiple targets is realized, and systematicness and overall benefits of decision making are improved.

Description

Technical Field

[0001] This invention relates to the field of dynamic pricing technology, and in particular to a dynamic pricing optimization method, system, and program product for the coal industry chain. Background Technology

[0002] Currently, pricing in the coal industry chain mainly relies on several technical solutions with inherent flaws: First, pricing based on manual experience through methods such as telephone inquiries and internal meetings, which is highly subjective and lacks standardization; second, pricing using simple rules such as cost-plus pricing and follow-the-leader pricing, whose rules are rigid and difficult to cope with complex and ever-changing markets; and third, using single models such as time series or linear regression for forecasting, which have limited dimensions of consideration and insufficient predictive accuracy. These existing methods generally suffer from significant problems of information acquisition and response lag. Due to the lack of automated data collection and processing mechanisms, it often takes several days from market changes to price adjustments, causing companies to miss the best pricing opportunities.

[0003] At the same time, these solutions also have the shortcoming of insufficient data utilization. They are unable to effectively integrate and connect the massive and multi-source data generated by various links in the supply, transportation and demand of the industrial chain, and lack the ability to fuse and analyze spatiotemporal data, resulting in a large amount of data value being buried.

[0004] Existing solutions have weak market response and forecasting capabilities. Due to the lack of quantitative models such as price elasticity analysis, they cannot accurately predict the impact of price changes on market supply and demand, often leading to inventory backlogs or stockouts. In addition, existing pricing strategies lack dynamic adaptability. Their evaluation systems are mostly static and cannot adaptively adjust according to changes in market conditions, resulting in pricing that is out of sync with actual market conditions. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a dynamic pricing optimization method, system and program product for the coal industry chain.

[0006] The present invention provides a dynamic pricing optimization method for the coal industry chain, comprising: The data acquisition and fusion module automatically acquires multi-source data from the coal industry chain and performs fusion processing on the acquired spatiotemporal data. The pricing forecasting module uses the fused data to predict market price trends and quantifies the response of price changes to supply and demand. The supply and demand analysis module uses the integrated data to predict and analyze the supply, demand, and inventory sides of the industrial chain. The quantitative evaluation module dynamically selects evaluation strategies based on the market environment and quantitatively scores the pricing scheme based on the results of the pricing forecast and supply and demand analysis. The optimization solution module uses a multi-objective optimization algorithm based on the quantitative scoring results to solve the problem and output the optimal pricing scheme.

[0007] In one of the optional technical solutions, the automatic acquisition of multi-source data from the coal industry chain through the data acquisition and fusion module, and the fusion processing of the acquired spatiotemporal data, includes: By connecting to coal trading platforms, logistics systems, meteorological services, and industry databases through application programming interfaces (APIs), it automatically collects pithead price data, transportation data, demand data, and market data. In the spatial dimension, a Gaussian kernel weighted average method based on geographic information system is used to aggregate the spatially dispersed price data of multiple pitheads into a regional benchmark price. In the time dimension, time series resampling technology is used to aggregate daily frequency data into weekly frequency data for predictive modeling.

[0008] In one of the optional technical solutions, the step of using a pricing prediction module to predict market price trends based on fused data and quantifying the response of price changes to supply and demand includes: A convergent model trainer framework is used to integrate Bayesian ridge regression and quantile regression models to predict market prices; We employ a dual elasticity decomposition strategy to quantitatively analyze the price response to supply and demand by decomposing it into two dimensions: value elasticity and sentiment elasticity.

[0009] In one of the optional technical solutions, the step of dynamically selecting an evaluation strategy based on the market environment using a quantitative evaluation module includes: During the supply-demand equilibrium period, the original assessment model is used for equilibrium assessment; During periods of sales pressure, switch to a sales-driven procurement evaluation model to prioritize sales. During periods of procurement difficulties, we switched to a procurement-based sales assessment model to prioritize cost control.

[0010] In one of the optional technical solutions, the quantitative scoring of the pricing scheme based on the pricing forecast and supply and demand analysis includes: Set hard constraints that must be met, and use a binary scoring mechanism for judgment; Set up soft constraints for the desired optimization, and use quadratic penalty function, piecewise function and linear function to calculate scores for scale indicators, interval indicators and ratio indicators respectively; Based on the scoring results of the hard and soft constraints, a comprehensive score for the pricing scheme is calculated by weighted average.

[0011] In one of the optional technical solutions, the optimization solution module, based on the quantitative scoring results, uses a multi-objective optimization algorithm to solve the problem and output the optimal pricing scheme, including: An improved genetic algorithm is used to solve the problem, and the improvement includes a constraint repair mechanism; The pricing decision is encoded as a real number vector, and the comprehensive score output by the quantitative evaluation module is used as the fitness function. The population is iteratively evolved through genetic operations of selection, crossover, and mutation until the convergence condition is met, and the pricing scheme with the highest comprehensive score is output.

[0012] The present invention provides a system for implementing any of the aforementioned dynamic pricing optimization methods in the coal industry chain, comprising: The data acquisition and fusion module is used to automatically acquire multi-source data from the coal industry chain and perform spatiotemporal data fusion. The pricing forecasting module is used to predict market price trends and quantify the supply and demand response to price changes; The supply and demand analysis module is used for forecasting and analyzing the supply, demand, and inventory sides of the industrial chain. The quantitative assessment module is used to dynamically select assessment strategies and quantitatively score pricing schemes. The optimization solution module is used to perform multi-objective optimization solutions based on the scoring results; The modules are connected sequentially.

[0013] The present invention provides an electronic device, including a memory, a processor, and an electronic device program on the memory, wherein the processor executes the electronic device program to implement the steps of any of the aforementioned dynamic pricing optimization methods for the coal industry chain.

[0014] The present invention provides an electronic device readable storage medium storing an electronic device program / instruction thereon, which, when executed by a processor, implements the steps of any of the aforementioned dynamic pricing optimization methods for the coal industry chain.

[0015] The present invention provides an electronic device program product, including an electronic device program / instruction, which, when executed by a processor, implements the steps of any of the aforementioned dynamic pricing optimization methods for the coal industry chain.

[0016] The above technical solution has the following beneficial effects: The dynamic pricing optimization method for the coal industry chain provided by this invention automatically integrates multi-source heterogeneous data through a data acquisition and fusion module, achieving real-time acquisition and deep integration of industry chain information and overcoming the problems of information lag and data silos. Through a pricing prediction and supply-demand analysis module, it accurately predicts market trends and quantifies price elasticity, enabling scientific judgment and response to market supply and demand relationships and solving the problem of weak prediction capabilities. Through a quantitative evaluation module, it dynamically selects evaluation strategies and scores schemes, achieving adaptive optimization of pricing strategies according to market conditions and breaking the rigidity of static evaluation. Finally, through an optimization solution module, it performs multi-objective global optimization to achieve collaborative optimal pricing while balancing multiple objectives, improving the systematic nature and overall efficiency of decision-making. Attached Figure Description

[0017] The disclosure of this invention will become more readily understood by referring to the accompanying drawings. It should be understood that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings: Figure 1 A flowchart illustrating a dynamic pricing optimization method for the coal industry chain provided in an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a dynamic pricing optimization system for the coal industry chain provided in an embodiment of the present invention; Figure 3 The execution flowchart of the pricing prediction module provided in an embodiment of the present invention is shown below; Figure 4 The execution flowchart of the supply and demand analysis module provided in one embodiment of the present invention; Figure 5 The execution flowchart of the quantitative evaluation module and the optimization solution module provided in an embodiment of the present invention is shown below; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0019] like Figure 1 The image shows a dynamic pricing optimization method for the coal industry chain provided by an embodiment of the present invention, comprising: Step S101: Automatically acquire multi-source data of the coal industry chain through the data acquisition and fusion module, and perform fusion processing on the acquired spatiotemporal data; Step S102: The pricing prediction module predicts market price trends based on the fused data and quantifies the response of price changes to supply and demand. Step S103: Based on the integrated data, the supply and demand analysis module performs predictive analysis on the supply side, demand side, and inventory side of the industrial chain; Step S104: The quantitative evaluation module dynamically selects an evaluation strategy based on the market environment and performs a quantitative score on the pricing scheme based on the results of the pricing forecast and supply and demand analysis. Step S105: Based on the quantitative scoring results, the optimization solution module uses a multi-objective optimization algorithm to solve the problem and output the optimal pricing scheme.

[0020] The overall technical architecture of this invention includes three layers: data layer, model layer, and decision layer, and achieves end-to-end automated pricing through five core functional modules.

[0021] First, at the data layer, step S101 forms the foundation of system operation, aiming to provide high-quality, standardized data input for subsequent model analysis and decision-making. The system, through an automated data acquisition and fusion module, obtains raw information such as pithead prices, transportation data, demand data, and market data in real time from multiple heterogeneous data sources, including coal trading platforms, logistics systems, meteorological services, and industry databases. Subsequently, the system employs spatiotemporal data fusion technology to aggregate these dispersed pithead prices with geographical location information into representative regional benchmark prices. Furthermore, it uses time-series resampling technology to aggregate daily data into weekly data for predictive modeling, forming a complete and consistent spatiotemporal dataset for the industrial chain.

[0022] Step S102 is responsible for analyzing the market from different dimensions. The pricing forecasting module receives the fused data and, on the one hand, uses a multi-model integration strategy to predict future market price trends. On the other hand, it uses an innovative elastic dual decomposition method to accurately quantify the impact of price changes on supply and demand, forming a price-quantity response model.

[0023] In step S103, the supply and demand analysis module, based on the same set of data, takes a more macro-level fundamental perspective to conduct comprehensive forecasting and analysis of the supply, demand, and inventory sides of the industrial chain, generating a supply and demand balance sheet and a market tension index to provide fundamental support for pricing decisions. The supply side includes domestic production and imports, the demand side includes various industries such as power, steel, chemicals, and building materials, and the inventory side includes ports, power plants, etc.

[0024] In step S104, the quantitative evaluation module integrates the results from the pricing forecasting module and the supply and demand analysis module. It does not employ fixed evaluation standards but dynamically selects the most suitable evaluation strategy based on the current market environment. On this basis, the quantitative evaluation module sets a series of hard and soft constraints, performing multi-dimensional and refined quantitative scoring on each candidate pricing scheme, calculating a comprehensive score, and thus transforming the decision quality into a comparable numerical value. Market environment conditions include supply and demand balance, sales pressure, and procurement difficulties; evaluation strategies include the original model, the sales-driven procurement model, and the procurement-driven sales model.

[0025] Step S105 generates the final pricing scheme based on the outputs of all the preceding steps. The optimization solution module uses the comprehensive score output by the quantitative evaluation module as the optimization objective, and employs multi-objective optimization techniques such as improved genetic algorithms to perform a global search under complex business constraints. Through iterative evolution, this module can automatically find the pricing scheme that achieves the best balance among multiple objectives such as profit, sales volume, and inventory, and finally outputs the specific purchase price of each station, the sales price of the port, and the predicted key business indicators, completing the entire process from data to decision.

[0026] In summary, the dynamic pricing optimization method for the coal industry chain provided by this invention automatically integrates multi-source heterogeneous data through a data acquisition and fusion module, achieving real-time acquisition and deep integration of industry chain information and overcoming the problems of information lag and data silos. Through a pricing prediction and supply-demand analysis module, it accurately predicts market trends and quantifies price elasticity, enabling scientific judgment and response to market supply and demand relationships and solving the problem of weak prediction capabilities. Through a quantitative evaluation module, it dynamically selects evaluation strategies and scores schemes, achieving adaptive optimization of pricing strategies according to market conditions and overcoming the rigidity of static evaluation. Finally, through an optimization solution module, it performs multi-objective global optimization, achieving collaborative optimal pricing while balancing multiple objectives, thus improving the systematic nature and overall efficiency of decision-making.

[0027] In one embodiment, step S101 includes the following sub-steps: By connecting to coal trading platforms, logistics systems, meteorological services, and industry databases through application programming interfaces (APIs), it automatically collects pithead price data, transportation data, demand data, and market data. In the spatial dimension, a Gaussian kernel weighted average method based on geographic information system is used to aggregate the spatially dispersed price data of multiple pitheads into a regional benchmark price. In the time dimension, time series resampling technology is used to aggregate daily frequency data into weekly frequency data for predictive modeling.

[0028] In this embodiment, the data acquisition and fusion module forms the data foundation of the system. Through standardized application programming interface (API) technology, the module establishes automated data connection channels with the coal trading platform, logistics management system, meteorological service platform, and industry-specific databases, enabling the automatic acquisition of pithead price data, railway and port transportation data, power plant demand data, and spot and futures market data.

[0029] In terms of spatial data processing, the system is based on a geographic information system platform and uses a Gaussian kernel function weighted average algorithm to calculate the spatial weight of each pithead around the benchmark station. The closer the pithead is to the benchmark station, the larger the weight coefficient. The smoothness of the Gaussian distribution is used to achieve continuous transition of spatial data, and finally a benchmark price that can represent the overall price level of the region is generated.

[0030] In terms of time dimension processing, the system adopts time series resampling technology to aggregate daily frequency data into weekly frequency data for predictive modeling, providing a consistent time benchmark for subsequent modeling and analysis.

[0031] In one embodiment, step S102 includes the following sub-steps: An aggregated model trainer framework is used to integrate Bayesian ridge regression, random sampling consensus algorithm and quantile regression model to predict market prices; We employ a dual elasticity decomposition strategy to quantitatively analyze the price response to supply and demand by decomposing it into two dimensions: value elasticity and sentiment elasticity.

[0032] In this embodiment, the pricing prediction module constructs a two-layer prediction analysis system.

[0033] In the market price prediction branch, the system uses an aggregate model trainer framework to train three core algorithm models in parallel: Bayesian Ridge Regression specifically addresses the multicollinearity problem among feature variables, providing stable coefficient estimates and a measure of prediction uncertainty; Random Sampling Consensus automatically identifies and eliminates outliers and noise interference in historical data, improving the robustness of the model; Quantile Regression provides price prediction intervals at different confidence levels, offering a reference for risk decision-making.

[0034] The predictions from the three models are integrated using a weighted average or stacked generalization strategy to ultimately output an accurate prediction of future market prices.

[0035] In the supply and demand response forecasting branch, the system innovatively proposes the elasticity dual decomposition theory, which decomposes the traditional single price elasticity into two independent dimensions: value elasticity and sentiment elasticity. A dual elasticity decomposition strategy is employed, using linear regression of historical data to estimate the value elasticity coefficient and the sentiment elasticity coefficient, respectively. Value elasticity reflects the impact of price competitiveness relative to the market benchmark, while sentiment elasticity reflects the impact of market psychological expectations and trends. The two elasticity coefficients work together to influence price changes, quantifying and predicting sales response.

[0036] Through refined analysis, the system can more accurately depict the actual impact of different pricing strategies on supply and demand quantities.

[0037] In one embodiment, step 104 includes the following sub-steps: During the supply-demand equilibrium period, the original assessment model is used for equilibrium assessment; During periods of sales pressure, switch to a sales-driven procurement evaluation model to prioritize sales. During periods of procurement difficulties, we switched to a procurement-based sales assessment model to prioritize cost control.

[0038] Furthermore, it also includes: Set hard constraints that must be met, and use a binary scoring mechanism for judgment; Set up soft constraints for the desired optimization, and use quadratic penalty function, piecewise function and linear function to calculate scores for scale indicators, interval indicators and ratio indicators respectively; Based on the scoring results of the hard and soft constraints, a comprehensive score for the pricing scheme is calculated by weighted average.

[0039] In this embodiment, the quantitative evaluation module establishes an intelligent dynamic evaluation mechanism. It dynamically selects an evaluation strategy based on the comparison between expected sales volume and expected shipment volume: when expected sales volume is less than expected shipment volume, it is determined that sales pressure is high, and the module switches to a sales-driven procurement evaluation mode; when expected sales volume is greater than or equal to expected shipment volume, it is determined that procurement pressure is high, and the module switches to a procurement-driven sales evaluation mode.

[0040] Furthermore, the quantitative evaluation module constructs a comprehensive scoring index system. The system categorizes evaluation constraints into two types: hard constraints and soft constraints. Hard constraints include bottom-line requirements that must be met, such as profits cannot be negative, transportation capacity cannot be exceeded, and prices must comply with policy restrictions. A binary scoring mechanism is used, with 1 point awarded for fully meeting all requirements and 0 points awarded for failing to meet any one of them.

[0041] Soft constraints include optimization indicators such as shipment volume targets, sales volume targets, and inventory control targets. A continuous scoring function is used to assign corresponding scores based on the degree of deviation between the actual value and the target value.

[0042] For different types of soft constraint indicators, the system has designed specialized scoring functions: for scale-based indicators such as sales volume, a quadratic penalty function is used, with more points deducted for larger deviations; for range-based indicators such as inventory levels, a piecewise function is used, with full marks awarded within a reasonable range and points deducted for exceeding the range; for ratio-based indicators such as cost percentage, a linear function is used, with scores calculated based on the proportional relationship.

[0043] The scores of each dimension are weighted and averaged according to preset weights to obtain the comprehensive score of the pricing scheme, which serves as the objective function for optimization.

[0044] In one embodiment, step S105 includes the following sub-steps: An improved genetic algorithm is used to solve the problem, and the improvement includes a constraint repair mechanism; The pricing decision is encoded as a real number vector, and the comprehensive score output by the quantitative evaluation module is used as the fitness function. The population is iteratively evolved through genetic operations of selection, crossover, and mutation until the convergence condition is met, and the pricing scheme with the highest comprehensive score is output.

[0045] In this embodiment, the optimization solution module employs an improved genetic algorithm for global optimization search. A constraint repair mechanism handles infeasible solutions generated during the evolution process through boundary mapping or penalty functions, ensuring the efficiency and feasibility of the search process.

[0046] The algorithm encodes pricing decision variables as real number vectors, with each vector representing a complete pricing scheme, and uses the comprehensive score output by the quantitative evaluation module as the fitness function.

[0047] Through genetic operations such as tournament selection, simulated binary crossover, and polynomial mutation, the quality of the population is continuously evolved until the convergence condition of the maximum number of iterations or continuous improvement of more than a threshold is met, and finally the optimal pricing scheme with the highest comprehensive score is output.

[0048] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0049] like Figure 2 As shown, the present invention provides a system for implementing any of the aforementioned dynamic pricing optimization methods in the coal industry chain, comprising: The data acquisition and fusion module is used to automatically acquire multi-source data from the coal industry chain and perform spatiotemporal data fusion. The pricing forecasting module is used to predict market price trends and quantify the supply and demand response to price changes; The supply and demand analysis module is used for forecasting and analyzing the supply, demand, and inventory sides of the industrial chain. The quantitative assessment module is used to dynamically select assessment strategies and quantitatively score pricing schemes. The optimization solution module is used to perform multi-objective optimization solutions based on the scoring results; The modules are connected sequentially.

[0050] Among them, such as Figure 3 The following is the execution flow of the pricing prediction module: The pricing forecasting module comprises two sub-modules that execute in parallel: market price forecasting and supply-demand response forecasting. The merged, standardized data is simultaneously input into both forecasting branches.

[0051] In the market price forecasting branch, the system first performs data preprocessing, using time series resampling technology to aggregate daily frequency data into weekly frequency data for forecasting modeling; then feature engineering is performed to construct lagged features to capture time series dependencies, construct cross features to capture the mutual influence between variables, and perform feature engineering processing, including lagged feature construction and differencing, to extract key features of the time series.

[0052] Next, the system trains three types of prediction models in parallel: Bayesian ridge regression specifically handles the multicollinearity problem among features and provides uncertainty estimation; random sampling consensus algorithm regression automatically identifies and eliminates the influence of outlier data points; and quantile regression provides prediction intervals at different confidence levels. The outputs of the three models are integrated through an aggregation model trainer framework. The system provides three integration strategies to choose from: simple averaging, weighted averaging, and stacked generalization. Finally, it outputs the market benchmark price and confidence interval for the next week.

[0053] In the supply and demand response forecasting branch, the system first performs benchmark point forecasting to determine future benchmark sales volume and benchmark price. Then, it employs an innovative dual elasticity decomposition strategy to decompose the price response into two dimensions: value elasticity and sentiment elasticity. Using this strategy, the value elasticity coefficient and sentiment elasticity coefficient are estimated separately through linear regression of historical data. Value elasticity reflects the competitive impact of price relative to the market benchmark, while sentiment elasticity reflects the impact of market psychological expectations and trends. These two elasticity coefficients work together to influence price changes, quantifying and predicting sales response.

[0054] The elasticity coefficient is estimated through linear regression of historical data, and the final response function can predict sales changes under different pricing strategies. The results of the two prediction branches are finally converged to form a complete pricing prediction output, which will serve as an important input to the quantitative evaluation module.

[0055] like Figure 4 The following is the execution flow of the supply and demand analysis module: The supply and demand analysis module integrates forecasting analysis across three dimensions—supply, demand, and inventory—to provide fundamental market support for pricing decisions.

[0056] The system first collects industrial economic data, electricity consumption data, and upstream and downstream industrial chain data from multiple data sources. After cleaning and standardization, this data is input into two parallel modules: supply-side forecasting and demand-side forecasting. Supply-side forecasting includes domestic production forecasting, transportation supply forecasting, and import supply forecasting. These three sub-modules work together to build a complete supply forecasting system. Demand-side forecasting is divided into two main categories: electricity demand and other industry demand. Electricity demand forecasting involves predicting total electricity consumption and then subtracting the generation of new energy sources such as hydropower, wind power, and photovoltaic power to obtain the thermal power generation demand, which is then converted into coal consumption. Other industry demand includes the raw material coal demand of the chemical industry, the coking coal demand of the metallurgical industry, and the fuel coal demand of the building materials industry. By modeling by industry, non-electricity coal demand is accurately grasped.

[0057] The forecasts from both the supply and demand sides converge in the inventory forecasting module. This module monitors the inventory index of the industrial chain, the inventory level of northern ports, and the inventory of key nodes such as power plants, and analyzes the supply and demand balance through inventory changes. The comprehensive analysis module performs domestic production and sales balance analysis and import and export impact assessment based on inventory data, and calculates the supply and demand tension index, ultimately generating a supply and demand balance sheet that includes market indices.

[0058] The entire supply and demand analysis system integrates more than forty subdivided prediction models and adopts a variety of modeling techniques, including the Prophet model for handling seasonal time series, the seasonal autoregressive integral moving average model for handling series with exogenous variables, and the extreme gradient boosting and lightweight gradient boosting machine algorithms for handling nonlinear relationships.

[0059] like Figure 5 The following is the execution flow of the quantitative evaluation module and the optimization solution module: The system first receives forecast results from the pricing forecast and supply and demand analysis modules, and then automatically selects an appropriate evaluation strategy based on the current market environment. When supply and demand are in balance, the original model is used for comprehensive and balanced assessment. When sales pressure is high, the model is switched to sales-driven procurement to ensure competitive sales prices. When procurement is difficult, the model is adopted to control costs and ensure profit margins.

[0060] After selecting the evaluation strategy, the system sets hard and soft constraints. Hard constraints include bottom-line requirements that must be met, such as non-negative profits, not exceeding transport capacity, and complying with policy price limits, and adopt a binary scoring mechanism. Soft constraints include optimization objectives such as shipment volume targets, sales volume targets, and inventory control, and adopt a continuous scoring function.

[0061] Design corresponding scoring functions for different types of indicators: Scale-related indicators use a quadratic penalty function to ensure that the greater the deviation, the heavier the penalty. Interval-related indicators use a piecewise function to ensure that full marks are obtained within a reasonable range. Proportion-related indicators use a linear function to reflect linear relationships. The scores of each dimension are calculated by weighted average to obtain a comprehensive score.

[0062] The comprehensive score is input as the objective function into the genetic algorithm optimization module. The algorithm first initializes the population, using real-number encoding to represent price decision variables. Then, it continuously improves the quality of solutions through evolutionary operations such as tournament selection to maintain population diversity, simulated binary crossover to achieve gene recombination, and polynomial mutation to introduce new features. Solutions that violate constraints are repaired through boundary mapping or penalty functions. The algorithm adaptively adjusts parameters according to the convergence stage. In the early stage, a high mutation rate is used to enhance global search capabilities; in the middle stage, exploration and development are balanced; and in the later stage, the mutation rate is reduced for refined searching. Through iterative optimization, new populations are continuously generated and re-evaluated until the convergence condition is met. Finally, an elite retention strategy is used to output the pricing scheme with the highest comprehensive score.

[0063] The execution flow of the data acquisition and fusion module is as follows: The data acquisition and fusion module connects to various data sources, including coal trading platforms, logistics systems, meteorological services, and industry databases, through application programming interfaces (APIs) to achieve automatic data acquisition.

[0064] The types of data collected include, but are not limited to: pithead price data containing geographical location, coal quality parameters, and production information; transportation data containing railway shipments, port throughput, and freight rates; demand data containing power plant daily consumption, inventory levels, and operating rates; and market data containing spot prices, futures prices, and price indices. Spatially, a Gaussian kernel-weighted average method based on a geographic information system is employed. Centered on a benchmark station, different weights are assigned to each pithead based on their geographical distance, with closer stations receiving larger weights. A Gaussian kernel function is used to achieve a smooth transition, thereby aggregating dispersed pithead prices into a regional benchmark price.

[0065] In the time dimension, time series resampling technology is used to aggregate daily frequency data into weekly frequency data for predictive modeling, and effective factors are screened through significance tests to provide a standardized and high-quality data foundation for subsequent analysis.

[0066] In summary, this invention, through the coordinated operation of the five core functional modules, constructs a complete decision-making closed loop, encompassing multi-source data acquisition, intelligent predictive analysis, dynamic quantitative evaluation, and global automatic optimization. This method effectively overcomes the inherent defects of traditional pricing technologies, such as information lag, data fragmentation, biased decision-making, and distorted predictions. It achieves intelligent, precise, and systematic pricing across the coal industry chain, significantly improving enterprises' market responsiveness and overall operational efficiency.

[0067] like Figure 6 The diagram shows a hardware structure of an electronic device according to the present invention, including a memory 602, a processor 601, and an electronic device program on the memory 602. The processor 601 executes the electronic device program to implement the steps of the dynamic pricing optimization method for the coal industry chain in any of the above embodiments.

[0068] Figure 6 Take the 601 processor as an example.

[0069] The electronic device may also include an input device 603 and a display device 604.

[0070] The processor 601, memory 602, input device 603 and display device 604 can be connected by a bus or other means. The figure shows an example of connection by a bus.

[0071] The memory 602, as a non-volatile electronic device readable storage medium, can be used to store non-volatile software programs, non-volatile electronic device executable programs, and modules, such as the program instructions / modules corresponding to the dynamic pricing optimization method for the coal industry chain in the embodiments of this application. The processor 601 executes various functional applications and data processing by running the non-volatile software programs, instructions, and modules stored in the memory 602, thereby realizing the dynamic pricing optimization method for the coal industry chain in the above embodiments.

[0072] The memory 602 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the dynamic pricing optimization method for the coal industry chain. Furthermore, the memory 602 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 602 may optionally include memory remotely located relative to the processor 601, and these remote memories can be connected via a network to the apparatus implementing the dynamic pricing optimization method for the coal industry chain. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0073] The input device 603 can receive user clicks and generate signal inputs related to user settings and function control of the dynamic pricing optimization method in the coal industry chain. The display device 604 may include display screens or other display devices.

[0074] When one or more modules are stored in the memory 602 and are run by one or more processors 601, the dynamic pricing optimization method for the coal industry chain in any of the above method embodiments is executed.

[0075] The electronic device disclosed in this invention, when in operation, can execute all the steps of the aforementioned dynamic pricing optimization method for the coal industry chain. Through a data acquisition and fusion module, it automatically integrates multi-source heterogeneous data, achieving real-time acquisition and deep fusion of industry chain information, overcoming the problems of information lag and data silos. Through a pricing prediction and supply-demand analysis module, it accurately predicts market trends and quantifies price elasticity, achieving scientific judgment and response to market supply and demand relationships, solving the problem of weak prediction capabilities. Through a quantitative evaluation module, it dynamically selects evaluation strategies and scores solutions, achieving adaptive optimization of pricing strategies according to market conditions, breaking the rigidity of static evaluation. Finally, through an optimization solution module, it performs multi-objective global optimization, achieving collaborative optimal pricing while balancing multiple objectives, improving the systematic nature and overall efficiency of decision-making.

[0076] An embodiment of the present invention provides an electronic device readable storage medium storing an electronic device program / instruction, which, when executed by a processor 601, implements all the steps of the dynamic pricing optimization method for the coal industry chain as described above.

[0077] In the context of this disclosure, a storage medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. The storage medium can be a machine-readable signal medium or a machine-readable storage medium. Optionally, the storage medium can be a non-transitory electronically readable storage medium, such as a ROM, random access memory (RAM), compact disc ROM (CD ROM), magnetic tape, floppy disk, and optical data storage device.

[0078] An embodiment of the present invention provides an electronic device program product, including an electronic device program / instruction, which, when executed by a processor, implements the steps of the dynamic pricing optimization method for the coal industry chain as described above.

[0079] By running the aforementioned electronic equipment program, all steps of the dynamic pricing optimization method for the coal industry chain described above can be executed. The data acquisition and fusion module automatically integrates multi-source heterogeneous data, achieving real-time acquisition and deep integration of industry chain information, overcoming information lag and data silos. The pricing prediction and supply-demand analysis module accurately predicts market trends and quantifies price elasticity, enabling scientific judgment and response to market supply and demand relationships, solving the problem of weak prediction capabilities. The quantitative evaluation module dynamically selects evaluation strategies and scores solutions, achieving adaptive optimization of pricing strategies according to market conditions, breaking the rigidity of static evaluation. Finally, the optimization solution module performs multi-objective global optimization, achieving synergistic optimal pricing while balancing multiple objectives, improving the systematic nature and overall efficiency of decision-making.

[0080] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A dynamic pricing optimization method for the coal industry chain, characterized in that, include: The data acquisition and fusion module automatically acquires multi-source data from the coal industry chain and performs fusion processing on the acquired spatiotemporal data. The pricing forecasting module uses the fused data to predict market price trends and quantifies the response of price changes to supply and demand. The supply and demand analysis module uses the integrated data to predict and analyze the supply, demand, and inventory sides of the industrial chain. The quantitative evaluation module dynamically selects evaluation strategies based on the market environment and quantitatively scores the pricing scheme based on the results of the pricing forecast and supply and demand analysis. The optimization solution module uses a multi-objective optimization algorithm based on the quantitative scoring results to solve the problem and output the optimal pricing scheme.

2. The dynamic pricing optimization method for the coal industry chain according to claim 1, characterized in that, The process of automatically acquiring multi-source data from the coal industry chain through a data acquisition and fusion module, and fusing the acquired spatiotemporal data, includes: By connecting to coal trading platforms, logistics systems, meteorological services, and industry databases through application programming interfaces (APIs), it automatically collects pithead price data, transportation data, demand data, and market data. In the spatial dimension, a Gaussian kernel weighted average method based on geographic information system is used to aggregate the spatially dispersed price data of multiple pitheads into a regional benchmark price. In the time dimension, time series resampling technology is used to aggregate daily frequency data into weekly frequency data for predictive modeling.

3. The dynamic pricing optimization method for the coal industry chain according to claim 1, characterized in that, The method of predicting market price trends based on fused data through a pricing prediction module, and quantifying the response of price changes to supply and demand, includes: A convergent model trainer framework is used to integrate Bayesian ridge regression and quantile regression models to predict market prices; We employ a dual elasticity decomposition strategy to quantitatively analyze the price response to supply and demand by decomposing it into two dimensions: value elasticity and sentiment elasticity.

4. The dynamic pricing optimization method for the coal industry chain according to claim 1, characterized in that, The process of dynamically selecting an evaluation strategy based on market conditions through a quantitative evaluation module includes: During the supply-demand equilibrium period, the original assessment model is used for equilibrium assessment; During periods of sales pressure, switch to a sales-driven procurement evaluation model to prioritize sales. During periods of procurement difficulties, we switched to a procurement-based sales assessment model to prioritize cost control.

5. The dynamic pricing optimization method for the coal industry chain according to claim 4, characterized in that, The quantitative scoring of the pricing scheme based on the pricing forecast and supply and demand analysis includes: Set hard constraints that must be met, and use a binary scoring mechanism for judgment; Set up soft constraints for the desired optimization, and use quadratic penalty function, piecewise function and linear function to calculate scores for scale indicators, interval indicators and ratio indicators respectively; Based on the scoring results of the hard and soft constraints, a comprehensive score for the pricing scheme is calculated by weighted average.

6. The dynamic pricing optimization method for the coal industry chain according to claim 1, characterized in that, The step of optimizing the solution module based on the quantitative scoring results, using a multi-objective optimization algorithm to solve the problem and output the optimal pricing scheme includes: An improved genetic algorithm is used to solve the problem, and the improvement includes a constraint repair mechanism; The pricing decision is encoded as a real number vector, and the comprehensive score output by the quantitative evaluation module is used as the fitness function. The population is iteratively evolved through genetic operations of selection, crossover, and mutation until the convergence condition is met, and the pricing scheme with the highest comprehensive score is output.

7. A system for implementing the dynamic pricing optimization method for the coal industry chain as described in any one of claims 1-6, characterized in that, include: The data acquisition and fusion module is used to automatically acquire multi-source data from the coal industry chain and perform spatiotemporal data fusion. The pricing forecasting module is used to predict market price trends and quantify the supply and demand response to price changes; The supply and demand analysis module is used for forecasting and analyzing the supply, demand, and inventory sides of the industrial chain. The quantitative assessment module is used to dynamically select assessment strategies and quantitatively score pricing schemes. The optimization solution module is used to perform multi-objective optimization solutions based on the scoring results; The modules are connected sequentially.

8. An electronic device, comprising a memory, a processor, and an electronic device program on the memory, characterized in that, The processor executes the electronic device program to implement the steps of the dynamic pricing optimization method for the coal industry chain as described in any one of claims 1-6.

9. An electronic device readable storage medium having an electronic device program / instructions stored thereon, characterized in that, When the electronic device program / instructions are executed by the processor, they implement the steps of the dynamic pricing optimization method for the coal industry chain as described in any one of claims 1-6.

10. An electronic device program product, comprising an electronic device program / instructions, characterized in that, When the electronic device program / instructions are executed by the processor, they implement the steps of the dynamic pricing optimization method for the coal industry chain as described in any one of claims 1-6.

Images