Pricing dynamic optimization system based on real-time detection of coal calorific value

Abstract

The application discloses a pricing dynamic optimization system based on real-time detection of coal calorific value, and relates to the field of coal product data management. The system comprises a data acquisition module, a data correlation module, a calorific value prediction module and a pricing optimization module. The data acquisition module is used for deploying LIBS online analyzers and microwave moisture meters on a coal production line to acquire coal production data streams in real time. The data correlation module is used for performing coal calorific value detection by adopting oxygen bomb calorimetry according to a preset acquisition period, acquiring coal calorific value data streams, and correlating and mapping the coal production data streams and the coal calorific value data streams. The calorific value prediction module is used for introducing boiler positive and negative balance data to perform calorific value prediction training and dynamic interference correction on a set based on coal calorific value data, and building a coal calorific value prediction channel. The pricing optimization module is used for performing pricing dynamic optimization on real-time coal calorific value data based on a dynamic pricing strategy. The technical problem of pricing deviation caused by untimely detection in the existing coal pricing system is solved, and the technical effect of coal pricing dynamic optimization is achieved.

Description

Technical Field

[0001] This application relates to the field of coal product data management, and in particular to a dynamic pricing optimization system based on real-time coal calorific value detection. Background Technology

[0002] In the coal trading market, coal calorific value is a key factor influencing coal prices, and its accuracy and real-time performance are crucial for ensuring fair trading and improving corporate economic efficiency. Currently, the main method to address the correlation between coal calorific value and pricing is to use periodic sampling and testing, basing pricing on coal calorific value data collected within a fixed period. However, this method, due to its fixed and long testing period, cannot reflect real-time changes in coal calorific value, leading to discrepancies between pricing and actual calorific value. This, in turn, affects the fairness of trading and corporate cost control.

[0003] Currently, in related technologies, coal pricing systems suffer from technical problems such as pricing deviations due to untimely detection. Summary of the Invention

[0004] This application provides a dynamic pricing optimization system based on real-time coal calorific value detection. It employs techniques such as deploying an online LIBS analyzer and a microwave moisture meter on the coal production line to collect production data in real time, obtaining calorific value data using the bomb calorimetry method at preset intervals and mapping it into a set of data pairs, introducing boiler forward and reverse balance data to train and correct the calorific value prediction model to build a prediction channel, and constructing a dynamic pricing strategy based on the real-time calorific value data output from the prediction channel to achieve pricing optimization. These techniques solve the technical problem of pricing deviations caused by untimely detection in existing coal pricing systems, and achieve the technical effect of dynamic optimization of coal pricing.

[0005] This application provides a dynamic pricing optimization system based on real-time coal calorific value detection, comprising: a data acquisition module for deploying an online LIBS analyzer and a microwave moisture meter on the coal production line, and acquiring coal production data streams in real time through the LIBS analyzer and microwave moisture meter; a data association module for performing coal calorific value detection using the bomb calorimetry method according to a preset acquisition cycle, acquiring coal calorific value data streams, and associating and mapping the coal production data streams and the coal calorific value data streams to establish a set of coal calorific value data pairs; a calorific value prediction module for introducing boiler forward and reverse balance data to perform calorific value prediction training and dynamic interference correction based on the set of coal calorific value data pairs, and establishing a coal calorific value prediction channel; and a pricing optimization module for outputting real-time coal calorific value data through the coal calorific value prediction channel, constructing a dynamic pricing strategy, and dynamically optimizing the pricing based on the real-time coal calorific value data based on the dynamic pricing strategy.

[0006] In a possible implementation, the data association module includes: a coal data distribution threshold determination module, used to determine a coal data distribution threshold according to coal data application standards; an abnormal data cleaning module, used to clean abnormal data from the coal production data stream and the coal calorific value data stream using the coal data distribution threshold, to obtain an available coal production data stream and an available coal calorific value data stream; a time-series arrangement module, used to arrange the available coal production data stream and the available coal calorific value data stream in time sequence, to obtain a coal production sequence data stream and a coal calorific value sequence data stream; and an association mapping module, used to perform time alignment and association mapping on the coal production sequence data stream and the coal calorific value sequence data stream, to establish the set of coal calorific value data pairs.

[0007] In a possible implementation, the calorific value prediction module includes: a set partitioning module, used to proportionally partition the coal calorific value data set to obtain a coal calorific value training set, a coal calorific value validation set, and a coal calorific value test set; a calorific value prediction training module, used to train the coal calorific value training set using a random forest network structure to generate an initial calorific value prediction channel; an iterative parameter tuning module, used to perform validation testing and evaluation and iterative parameter tuning on the initial calorific value prediction channel using the coal calorific value validation set and the coal calorific value test set to obtain a basic calorific value prediction channel; and a dynamic interference correction module, used to introduce boiler forward and reverse balance data to dynamically correct the interference of the basic calorific value prediction channel to obtain a coal calorific value prediction channel.

[0008] In a possible implementation, the dynamic interference correction module includes: a positive balance efficiency calculation module, used to extract input and output heat data and calculate efficiency ratios from the boiler positive and negative balance data to obtain boiler positive balance efficiency data; an anti-balance efficiency calculation module, used to extract heat loss data and calculate efficiency loss from the boiler positive and negative balance data to obtain boiler anti-balance efficiency data; a deviation comparison module, used to compare the deviations of the boiler positive balance efficiency data and the boiler anti-balance efficiency data to obtain positive and negative balance efficiency deviation parameters; and a channel correction module, used to dynamically correct the basic calorific value prediction channel based on the positive and negative balance efficiency deviation parameters to obtain a coal calorific value prediction channel.

[0009] In a possible implementation, the channel correction module includes: a deviation interference threshold setting module, used to set a deviation interference threshold based on boiler design parameters and actual operating experience; a correction fitting module, used to perform correction fitting based on historical boiler operating data and establish a balance deviation parameter-calorific value correction coefficient function if the positive and negative balance efficiency deviation parameter is greater than the deviation interference threshold; a dynamic calculation module, used to dynamically calculate the positive and negative balance efficiency deviation parameter using the balance deviation parameter-calorific value correction coefficient function to determine the target calorific value correction coefficient; and an interference correction module, used to perform dynamic interference correction on the basic calorific value prediction channel based on the target calorific value correction coefficient to obtain the coal calorific value prediction channel.

[0010] In a possible implementation, the dynamic calculation module includes: an absolute value calculation module, used to dynamically calculate the absolute value of the positive and negative balance efficiency deviation parameter using the balance deviation parameter-calorific value correction coefficient function to obtain a scalar calorific value correction coefficient; a correction direction determination module, used to determine the calorific value correction direction based on the positive and negative balance efficiency deviation parameter; and a vector marking module, used to vector mark the scalar calorific value correction coefficient according to the calorific value correction direction to determine the target calorific value correction coefficient.

[0011] In a possible implementation, the pricing optimization module includes: a calorific value comparison module, used to obtain the contract benchmark calorific value, compare the predicted calorific value of coal with the contract benchmark calorific value, and obtain the coal calorific value comparison result; and a dynamic pricing strategy acquisition module, used to perform historical coal pricing data mining and piecewise linear function fitting based on the coal calorific value comparison result, to obtain a dynamic pricing strategy.

[0012] In a possible implementation, the dynamic pricing strategy acquisition module includes: a historical coal pricing data mining module, used to perform historical coal pricing data mining based on the coal calorific value comparison results to obtain a first coal pricing dataset and a second coal pricing dataset, wherein the first coal pricing dataset is a historical pricing dataset where the real-time calorific value of coal is greater than the contract benchmark calorific value, and the second coal pricing dataset is a historical pricing dataset where the real-time calorific value of coal is less than the contract benchmark calorific value; and a piecewise linear function fitting module, used to perform piecewise linear function fitting using the first coal pricing dataset and the second coal pricing dataset respectively to determine the dynamic pricing strategy.

[0013] In a possible implementation, the piecewise linear function fitting module includes: a coal profit objective function construction module for constructing a coal profit objective function; an evaluation and screening module for evaluating and screening the first coal pricing dataset and the second coal pricing dataset using the coal profit objective function to obtain a first pricing optimization dataset and a second pricing optimization dataset; and a fitting module for performing piecewise linear function fitting based on the first pricing optimization dataset and the second pricing optimization dataset to determine the dynamic pricing strategy.

[0014] In a possible implementation, the fitting module includes: a first fitting module, used to perform piecewise linear function fitting based on the first pricing optimization dataset to obtain a first dynamic function for calorific value pricing; a second fitting module, used to perform piecewise linear function fitting based on the second pricing optimization dataset to obtain a second dynamic function for calorific value pricing; and a result determination module, used to determine the dynamic pricing strategy based on the first dynamic function for calorific value pricing and the second dynamic function for calorific value pricing.

[0015] This application proposes a dynamic pricing optimization system based on real-time coal calorific value detection. This system deploys an online LIBS analyzer and a microwave moisture meter on the coal production line via a data acquisition module. The LIBS analyzer and microwave moisture meter acquire real-time coal production data streams. A data association module performs coal calorific value detection using bomb calorimetry according to a preset acquisition cycle, acquiring a coal calorific value data stream. The coal production data stream and the coal calorific value data stream are correlated and mapped to establish a set of coal calorific value data pairs. A calorific value prediction module incorporates boiler forward and reverse balance data and performs calorific value prediction training and dynamic interference correction based on the set of coal calorific value data pairs, establishing a coal calorific value prediction channel. A pricing optimization module outputs real-time coal calorific value data through the coal calorific value prediction channel and constructs a dynamic pricing strategy. Based on this dynamic pricing strategy, the system dynamically optimizes the pricing of the coal based on the real-time calorific value data. This achieves the technical effect of dynamic optimization of coal pricing. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Structural diagrams are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0017] Figure 1 This is a schematic diagram of the pricing dynamic optimization system based on real-time detection of coal calorific value provided in an embodiment of this application.

[0018] Figure 2 This is a schematic diagram of the calorific value prediction module in the pricing dynamic optimization system based on real-time coal calorific value detection provided in this application embodiment.

[0019] Figure 3 This is a schematic diagram of the dynamic interference correction module in the pricing dynamic optimization system based on real-time coal calorific value detection provided in an embodiment of this application.

[0020] Figure labeling: Data acquisition module 10, data association module 20, calorific value prediction module 30, pricing optimization module 40. Detailed Implementation

[0021] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below.

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0024] This application provides a pricing dynamic optimization system based on real-time detection of coal calorific value, such as... Figure 1 As shown, the system includes: The data acquisition module 10 is used to deploy an online LIBS analyzer and a microwave moisture meter on the coal production line, and to acquire coal production data streams in real time through the online LIBS analyzer and the microwave moisture meter.

[0025] Specifically, LIBS (Laser-Induced Breakdown Spectroscopy) online analyzers utilize a high-energy pulsed laser beam focused onto the surface of a coal sample, generating high-temperature plasma in a localized area. Atoms and ions in the plasma emit characteristic spectra during cooling. These spectral signals are collected by a spectrometer, and the chemical composition of the coal is analyzed based on these spectral characteristics. For example, the content of elements such as carbon, hydrogen, oxygen, and sulfur in coal can be detected. Microwave moisture analyzers measure the moisture content of coal based on the differences in the propagation characteristics of microwaves in different media. When microwaves pass through coal, parameters such as the phase and amplitude of the microwaves change; by measuring these changes, the moisture content of the coal can be accurately calculated.

[0026] Specifically, an online LIBS analyzer and a microwave moisture meter are installed on the coal production line. The laser emitter of the LIBS analyzer is aimed at the coal sample on the coal conveyor belt, and the microwave emitter and receiver of the microwave moisture meter are also installed in appropriate locations. When the equipment is turned on, the LIBS analyzer emits laser light at a set sampling frequency (e.g., once per second), collects the spectral signal, and performs preliminary processing. Simultaneously, the microwave moisture meter emits microwave signals, receives the reflected signals, and calculates the moisture content. The collected coal chemical composition data and moisture content data are packaged into a coal production data stream and sent to the data association module 20 via a data transmission network (e.g., industrial Ethernet).

[0027] The data association module 20 is used to perform coal calorific value detection using the oxygen bomb calorimetry method according to a preset acquisition cycle, acquire coal calorific value data stream, associate and map the coal production data stream and the coal calorific value data stream, and establish a set of coal calorific value data pairs.

[0028] Specifically, the bomb calorimetry method is a method for measuring the calorific value of coal. A certain amount of coal sample is placed in an oxygen bomb, filled with oxygen, and burned in a calorimeter. The calorific value of the coal is determined by measuring the heat released during combustion. Samples are periodically taken from the coal production line (according to a preset collection cycle, such as once per hour) and their calorific value is measured using a bomb calorimeter, resulting in a coal calorific value data stream. The calorific value data obtained by the bomb calorimetry method is stored in a database, along with the coal production data stream sent by the data acquisition module 10. A data association algorithm is used to associate the coal production data stream and the coal calorific value data stream according to time sequence or batch number. For example, if a record in the coal production data stream has a timestamp of "2025-06-11 10:00:00", and a record with a corresponding timestamp exists in the calorific value data obtained by the bomb calorimetry method, these two data sets are associated to form a set of coal calorific value data pairs.

[0029] In one possible implementation, the data association module 20 includes: a coal data distribution threshold determination module, used to determine a coal data distribution threshold according to coal data application standards; an abnormal data cleaning module, used to clean abnormal data from the coal production data stream and the coal calorific value data stream using the coal data distribution threshold to obtain an available coal production data stream and an available coal calorific value data stream; a time-series arrangement module, used to arrange the available coal production data stream and the available coal calorific value data stream in time sequence to obtain a coal production sequence data stream and a coal calorific value sequence data stream; and an association mapping module, used to perform time alignment and association mapping on the coal production sequence data stream and the coal calorific value sequence data stream to establish the set of coal calorific value data pairs.

[0030] Specifically, historical coal production and calorific value data are collected, including chemical composition, moisture content, and calorific value. By analyzing a large amount of coal production and calorific value data, statistical parameters such as the mean and standard deviation are calculated. For example, for coal ash content data, the average value is calculated to be 15%, and the standard deviation to be 2%. Based on the distribution pattern of the coal data (such as normal or skewed distribution), a data distribution threshold is determined. For example, for ash content data with an approximately normal distribution, the threshold can be set to the mean plus or minus two standard deviations, i.e., 9%-21%.

[0031] Each data point in the coal production and coal calorific value data streams is compared to its corresponding distribution threshold. If a data point exceeds the threshold, it is marked as outlier and deleted. For example, a data point with an ash content of 22% is deleted from the data stream. Edge data (data points close to the threshold) is processed using a data smoothing algorithm (such as a moving average) to reduce the impact of data fluctuations on the results. For example, for data with an ash content of 20.5%, a moving average is used to calculate its smoothed value.

[0032] Extract timestamp information from the available coal production data stream and the available coal calorific value data stream. Sort the data using a timestamp sorting algorithm to obtain a data stream arranged in chronological order. Divide the sorted data into different time windows. For example, divide the data into hourly time windows to obtain the coal production sequence data stream and the coal calorific value sequence data stream.

[0033] A time alignment algorithm is employed to align the coal production sequence data stream and the coal calorific value sequence data stream in time. For example, an interpolation algorithm is used to align the data in time. The time-aligned data is then correlated and mapped to establish a set of coal calorific value data pairs. For instance, coal production data {timestamp: "2025-06-11 10:00:00", carbon content: 75%, hydrogen content: 5%, moisture content: 10%} and calorific value data {timestamp: "2025-06-11 10:00:00", calorific value: 6000 kcal / kg} are correlated to form a set of data pairs. This implementation, through a coal data distribution threshold determination module and an anomaly data cleaning module, can effectively identify and clean abnormal data, improving the accuracy and reliability of the data. The time sequence arrangement module and the correlation mapping module can arrange and correlate the coal production data stream and the coal calorific value data stream in chronological order, establishing a more accurate set of coal calorific value data pairs. Based on a more accurate and reliable set of coal calorific value data, the pricing optimization module 40 can construct a more reasonable dynamic pricing strategy and improve the efficiency of enterprises.

[0034] The calorific value prediction module 30 is used to introduce boiler positive and negative balance data, perform calorific value prediction training and dynamic interference correction on the set based on the coal calorific value data, and build a coal calorific value prediction channel.

[0035] Specifically, a set of coal calorific value data pairs is used as the training dataset, and boiler forward and reverse balance data (forward balance data calculates boiler thermal efficiency based on fuel consumption, air volume, etc., while reverse balance data calculates thermal efficiency based on boiler flue gas losses, incomplete combustion losses, etc.) are introduced as auxiliary features to train the calorific value prediction model. By monitoring changes in boiler operating status (such as load changes, combustion efficiency changes, etc.), the parameters of the calorific value prediction model are adjusted in real time. For example, when a sudden increase in boiler load is detected, the calorific value prediction results are corrected according to pre-set disturbance correction rules.

[0036] Specifically, a large amount of coal calorific value data pairs and boiler forward and reverse balance data are collected, and the data is preprocessed (e.g., normalization, outlier removal). A suitable machine learning algorithm is selected, and the processed data is input into the algorithm for training. For example, using a neural network algorithm, the number of neurons in the input layer is set to the sum of the number of features in the coal calorific value data pairs and the number of features in the boiler forward and reverse balance data; the number of neurons in the hidden layer is determined empirically or through cross-validation; and the number of neurons in the output layer is 1 (predicted calorific value). After the model is trained, it is deployed to the coal calorific value prediction channel. Simultaneously, a dynamic disturbance correction system is activated to monitor changes in the boiler's operating status in real time. When disturbance is detected, the parameters of the calorific value prediction model are adjusted according to the type and degree of disturbance, thereby achieving dynamic disturbance correction for calorific value prediction.

[0037] like Figure 2 As shown, in one possible implementation, the calorific value prediction module 30 includes: a set partitioning module, used to proportionally partition the coal calorific value data set to obtain a coal calorific value training set, a coal calorific value validation set, and a coal calorific value test set; a calorific value prediction training module, used to train the coal calorific value training set using a random forest network structure to generate an initial calorific value prediction channel; an iterative parameter tuning module, used to perform verification testing and evaluation and iterative parameter tuning on the initial calorific value prediction channel using the coal calorific value validation set and the coal calorific value test set to obtain a basic calorific value prediction channel; and a dynamic interference correction module, used to introduce boiler forward and reverse balance data to dynamically correct the interference of the basic calorific value prediction channel to obtain a coal calorific value prediction channel.

[0038] Specifically, data partitioning methods can be employed, such as dividing the coal calorific value data set into training, validation, and test sets according to a fixed ratio (e.g., 70%-15%-15%). Random partitioning or partitioning according to time series can be used to ensure data representativeness. For example, a random number generator can be used to randomly shuffle the coal calorific value data set to ensure data randomness. Data can also be partitioned according to a set ratio. For example, if the coal calorific value data set has 1000 samples, the training set consists of the first 700 samples, the validation set consists of the middle 150 samples, and the test set consists of the last 150 samples. Alternatively, partitioning can be done according to time series, for example, using the first 70% of the time window data as the training set, the middle 15% as the validation set, and the last 15% as the test set.

[0039] Random Forest is an ensemble learning algorithm that improves the accuracy and stability of predictions by constructing multiple decision trees. Each decision tree is trained using randomly sampled data and feature subsets. The coal calorific value training data is preprocessed, such as by normalization and removal of missing values. Parameters of the random forest are set, such as the number of decision trees (e.g., 100 trees), maximum depth, and minimum number of splits. The random forest model is trained using the coal calorific value training data, for example, by selecting the optimal parameter combination through cross-validation. After training, an initial calorific value prediction channel is obtained, which can be used to predict the calorific value of coal.

[0040] The initial calorific value prediction channel was evaluated using a coal calorific value validation set, and performance metrics (such as mean squared error and accuracy) were calculated. Based on the evaluation results, the parameters of the random forest (such as the number of decision trees and maximum depth) were adjusted. The model was retrained using the adjusted parameters, and the above evaluation and adjustment process was repeated until the performance metrics no longer showed significant improvement. Finally, the optimized model was evaluated using a coal calorific value test set to ensure the model's generalization ability.

[0041] Boiler forward and reverse balance data are collected, including boiler load, combustion efficiency, and flue gas temperature. This data is used as an auxiliary feature to design dynamic disturbance correction rules for the basic calorific value prediction channel. For example, when the boiler load suddenly increases, the calorific value prediction result is appropriately increased; when the combustion efficiency decreases, the calorific value prediction result is appropriately decreased. These correction rules are integrated into the basic calorific value prediction channel, and the boiler operating status is monitored in real time, with the calorific value prediction result dynamically adjusted according to changes in status. In this implementation, the dynamic disturbance correction module incorporates boiler forward and reverse balance data, enabling dynamic adjustment of the calorific value prediction result based on the boiler operating status. This allows the model to better adapt to dynamic changes in actual production, improving the practicality and reliability of the prediction results.

[0042] like Figure 3 As shown, in one possible implementation, the dynamic interference correction module includes: a positive balance efficiency calculation module, used to extract input and output heat data and calculate efficiency ratios from the boiler positive and negative balance data to obtain boiler positive balance efficiency data; an anti-balance efficiency calculation module, used to extract heat loss data and calculate efficiency loss from the boiler positive and negative balance data to obtain boiler anti-balance efficiency data; a deviation comparison module, used to compare the deviations of the boiler positive balance efficiency data and the boiler anti-balance efficiency data to obtain positive and negative balance efficiency deviation parameters; and a channel correction module, used to dynamically correct the basic calorific value prediction channel based on the positive and negative balance efficiency deviation parameters to obtain a coal calorific value prediction channel.

[0043] Specifically, the input heat (such as fuel calorific value, air volume, etc.) and output heat (such as the heat of steam generated by the boiler) are extracted from the boiler's positive and negative balance data. The ratio of input heat to output heat is calculated to obtain the boiler's positive balance efficiency. For example, the positive balance efficiency can be calculated using the following formula: Positive balance efficiency = Output heat / Input heat × 100%.

[0044] Heat loss data, such as flue gas loss, chemical incomplete combustion loss, and mechanical incomplete combustion loss, is extracted from the boiler's forward and reverse balance data. The percentage of each type of heat loss relative to the total input heat is calculated to obtain the boiler's reverse balance efficiency. For example, the reverse balance efficiency can be calculated using the following formula: Reverse balance efficiency = 100% − (Flue gas loss + Chemical incomplete combustion loss + Mechanical incomplete combustion loss).

[0045] The deviation between the boiler's positive and negative balance efficiency data is calculated. For example, the deviation can be calculated using the formula: Positive / Negative Balance Efficiency Deviation = Positive Balance Efficiency - Negative Balance Efficiency. The magnitude and direction of the deviation are analyzed to determine if the boiler's operating status is normal. Based on the positive / negative balance efficiency deviation parameters, the parameters of the calorific value prediction model are adjusted. For example, if the deviation is positive, the calorific value prediction result is appropriately increased; if the deviation is negative, the calorific value prediction result is appropriately decreased. The adjusted model parameters are then updated in the basic calorific value prediction channel. The boiler's operating status is monitored in real time, and the coal calorific value prediction channel is dynamically adjusted based on the new deviation parameters. In this implementation, the channel correction module dynamically adjusts the calorific value prediction channel based on the positive / negative balance efficiency deviation parameters, enabling it to adapt to changes in the boiler's operating status in real time. This dynamic correction mechanism allows the calorific value prediction model to not only handle static data but also cope with dynamic changes in actual production, improving the model's adaptability and robustness.

[0046] For example, suppose the boiler input heat is 10000 MJ and the output heat is 8000 MJ. Positive balance efficiency = 8000 / 10000 × 100% = 80%. Assume the boiler's heat loss data is: flue gas loss 10%, chemical incomplete combustion loss 5%, and mechanical incomplete combustion loss 3%. Reverse balance efficiency = 100% − (10% + 5% + 3%) = 82%. Positive-reverse balance efficiency deviation = 80% − 82% = −2%. A negative deviation indicates that the reverse balance efficiency is higher than the positive balance efficiency, suggesting that the boiler's actual operating efficiency is lower than expected, and the actual calorific value of the coal may be overestimated. Based on the deviation parameter (-2%), adjust the basic calorific value prediction channel to make it closer to the actual operating state.

[0047] In one possible implementation, the channel correction module includes: a deviation interference threshold setting module, used to set a deviation interference threshold based on boiler design parameters and actual operating experience; a correction fitting module, used to perform correction fitting based on historical boiler operating data and establish a balance deviation parameter-calorific value correction coefficient function if the positive and negative balance efficiency deviation parameter is greater than the deviation interference threshold; a dynamic calculation module, used to dynamically calculate the positive and negative balance efficiency deviation parameter using the balance deviation parameter-calorific value correction coefficient function to determine the target calorific value correction coefficient; and an interference correction module, used to perform dynamic interference correction on the basic calorific value prediction channel based on the target calorific value correction coefficient to obtain the coal calorific value prediction channel.

[0048] Specifically, based on boiler design parameters (such as rated efficiency and design heat loss) and actual operating experience (such as historical operating data and common deviation ranges), a reasonable deviation interference threshold should be set. For example, analyze the boiler design parameters to determine the theoretical range of forward and reverse balance efficiency deviations. Combine historical operating data to statistically analyze the distribution of forward and reverse balance efficiency deviations in actual operation. Integrate the design parameters and historical data to set a reasonable deviation interference threshold, such as ±2%.

[0049] If the positive and negative balance efficiency deviation parameter exceeds the deviation interference threshold, it indicates a significant deviation between the current boiler operating state and the normal state, requiring correction of the calorific value prediction model. Based on historical boiler operating data, a data fitting algorithm (such as linear regression, polynomial fitting, etc.) is used to establish a functional relationship between the balance deviation parameter and the calorific value correction coefficient. For example, the positive and negative balance efficiency deviation parameter and the corresponding calorific value correction coefficient are extracted from historical boiler operating data, and a data fitting algorithm (such as linear regression) is used to fit the data to establish a functional relationship. For example, assuming the fitting result is: calorific value correction coefficient = a × positive and negative balance efficiency deviation + b, where a and b are fitting parameters, the fitted function is stored as the balance deviation parameter - calorific value correction coefficient function.

[0050] The target calorific value correction coefficient is calculated by substituting the current forward and reverse balance efficiency deviation parameters into the balance deviation parameter-calorific value correction coefficient function. For example, if the current forward and reverse balance efficiency deviation is -3%, the target calorific value correction coefficient is calculated as: target calorific value correction coefficient = a × (−3%) + b, and the result is used as the target calorific value correction coefficient. Based on the target calorific value correction coefficient, the basic calorific value prediction channel is dynamically corrected to obtain the final coal calorific value prediction channel. In this implementation, the deviation interference threshold setting module reasonably sets the deviation interference threshold by combining boiler design parameters and actual operating experience, ensuring that correction is only performed when there is a significant deviation between the actual operating state and the normal state, avoiding unnecessary adjustments. The correction fitting module establishes a functional relationship between the balance deviation parameter and the calorific value correction coefficient based on historical data, which can accurately calculate the appropriate correction coefficient according to the current deviation, improving the scientific nature and accuracy of the correction.

[0051] In one possible implementation, the dynamic calculation module includes: an absolute value calculation module, used to dynamically calculate the absolute value of the positive and negative balance efficiency deviation parameter using the balance deviation parameter-calorific value correction coefficient function to obtain a scalar calorific value correction coefficient; a correction direction determination module, used to determine the calorific value correction direction based on the positive and negative balance efficiency deviation parameter; and a vector marking module, used to vector mark the scalar calorific value correction coefficient according to the calorific value correction direction to determine the target calorific value correction coefficient.

[0052] Specifically, the absolute value of the positive and negative balance efficiency deviation parameters is calculated using the balance deviation parameter-calorific value correction coefficient function to obtain a scalar calorific value correction coefficient. The magnitude of the deviation (regardless of direction) is converted into a scalar value of the correction coefficient. The direction of calorific value correction is determined based on the sign of the positive and negative balance efficiency deviation parameters. A positive deviation indicates that the calorific value correction coefficient needs to be increased, and a negative deviation indicates that the calorific value correction coefficient needs to be decreased. Based on the correction direction, the scalar calorific value correction coefficient is converted into a vector form, i.e., the target calorific value correction coefficient. If the correction direction is increasing, the target calorific value correction coefficient is the scalar calorific value correction coefficient. If the correction direction is decreasing, the target calorific value correction coefficient is the reciprocal of the scalar calorific value correction coefficient. In this implementation, the absolute value calculation module ensures that the magnitude of the correction coefficient depends only on the magnitude of the deviation and is independent of the direction of the deviation by calculating the absolute value of the positive and negative balance efficiency deviation parameters, making the calculation of the correction coefficient more scientific and accurate. The correction direction determination module determines the direction of correction by judging the sign of the deviation, ensuring that the direction of correction is consistent with the actual operating state, which further improves the accuracy of the correction.

[0053] The pricing optimization module 40 is used to output real-time coal calorific value data through the coal calorific value prediction channel, construct a dynamic pricing strategy, and dynamically optimize the pricing of the real-time coal calorific value data based on the dynamic pricing strategy.

[0054] Specifically, the system receives real-time coal calorific value data from the coal calorific value prediction channel and uses it as input to a dynamic pricing model. Based on market research data (such as coal market price fluctuations and supply-demand relationships), the parameters of the dynamic pricing model are set. For example, the price might increase by a certain percentage for every 1000 kcal increase in calorific value. An optimization algorithm is then initiated to optimize the dynamic pricing model. For instance, a genetic algorithm is used to initialize a population (each population representing a combination of pricing parameters), and through selection, crossover, and mutation operations, the optimal combination of pricing parameters is found after multiple generations of iteration. Based on the optimized dynamic pricing strategy, the real-time coal calorific value data is dynamically optimized for pricing, and the optimized price is sent to the sales system or displayed on a price display screen.

[0055] In one possible implementation, the pricing optimization module 40 includes: a calorific value comparison module, used to obtain the contract benchmark calorific value, compare the predicted calorific value of coal with the contract benchmark calorific value, and obtain the coal calorific value comparison result; and a dynamic pricing strategy acquisition module, used to perform historical coal pricing data mining and piecewise linear function fitting based on the coal calorific value comparison result, and obtain a dynamic pricing strategy.

[0056] Specifically, the predicted calorific value of coal and the contract benchmark calorific value are obtained from the data source. The predicted calorific value of coal is compared with the contract benchmark calorific value, and the difference between the two is calculated. For example: calorific value deviation = predicted calorific value of coal - contract benchmark calorific value. The calorific value deviation and its direction (positive or negative deviation) are output as the comparison result.

[0057] Historical coal pricing data, including calorific value deviation and corresponding price adjustments, is extracted from the database. Based on this data, a piecewise linear function fitting method is used to establish the relationship between calorific value deviation and coal price. For example, the calorific value deviation can be divided into multiple intervals, each corresponding to a linear function. For instance: Interval 1: calorific value deviation ≤ -5%, Interval 2: -5% < calorific value deviation ≤ 0%, Interval 3: 0% < calorific value deviation ≤ 5%, Interval 4: calorific value deviation > 5%. Linear regression is used to fit the relationship between calorific value deviation and price adjustment for each interval, and the fitted piecewise linear function is used as a dynamic pricing strategy. In this implementation, the calorific value comparison module accurately calculates the calorific value deviation by comparing the predicted calorific value of coal with the contract benchmark calorific value, providing a scientific basis for the pricing strategy. The dynamic pricing strategy acquisition module, through historical data mining and piecewise linear function fitting, can dynamically adjust coal prices based on calorific value deviation, ensuring more accurate and reasonable pricing.

[0058] In one possible implementation, the dynamic pricing strategy acquisition module includes: a historical coal pricing data mining module, used to perform historical coal pricing data mining based on the coal calorific value comparison results to obtain a first coal pricing dataset and a second coal pricing dataset, wherein the first coal pricing dataset is a historical pricing dataset where the real-time calorific value of coal is greater than the contract benchmark calorific value, and the second coal pricing dataset is a historical pricing dataset where the real-time calorific value of coal is less than the contract benchmark calorific value; and a piecewise linear function fitting module, used to perform piecewise linear function fitting using the first coal pricing dataset and the second coal pricing dataset respectively to determine the dynamic pricing strategy.

[0059] Specifically, all coal pricing records are extracted from the historical database, including information such as coal calorific value, contract benchmark calorific value, and price adjustments. Based on the coal calorific value comparison results, the historical coal pricing data is divided into two datasets: the first dataset contains historical pricing data where the real-time coal calorific value is greater than the contract benchmark calorific value; the second dataset contains historical pricing data where the real-time coal calorific value is less than the contract benchmark calorific value. That is, the first dataset has a positive calorific value deviation, and the second dataset has a negative calorific value deviation. Each dataset is cleaned and preprocessed to remove outliers and missing values.

[0060] Piecewise linear function fitting was performed on the first and second coal pricing datasets respectively. Piecewise linear regression was used to capture pricing patterns within different calorific value deviation intervals. Specifically, based on the distribution of calorific value deviation, the data was divided into multiple intervals, and a linear function was used to fit each interval. Specifically, the first coal pricing dataset was analyzed to determine the interval divisions for calorific value deviation. For example: Interval 1: calorific value deviation > 0% and ≤ 5%; Interval 2: calorific value deviation > 5% and ≤ 10%; Interval 3: calorific value deviation > 10%. Linear regression was used to fit the relationship between calorific value deviation and price adjustment for each interval. For example, for interval 1, the fitting function was: Price Adjustment = x1 × Calorific Value Deviation + y1, where x1 and y1 are the coefficients of the linear regression model.

[0061] A similar analysis was performed on the second coal pricing dataset to determine the interval divisions for calorific value deviation. For example: Interval 1: calorific value deviation < 0% and ≥ -5%; Interval 2: calorific value deviation < -5% and ≥ -10%; Interval 3: calorific value deviation < -10%. Linear regression was used to fit the relationship between calorific value deviation and price adjustment for each interval. For example, for interval 1, the fitted function was: Price Adjustment = x2 × Calorific Value Deviation + y2, where x2 and y2 are the coefficients of the linear regression model. The fitted piecewise linear functions were integrated into a dynamic pricing strategy. In this implementation, the historical coal pricing data mining module, by dividing historical data into two datasets to handle cases where the calorific value is higher than and lower than the contract benchmark, can more accurately capture pricing patterns under different conditions. The piecewise linear function fitting module, through piecewise linear regression, can better adapt to the changing patterns of calorific value deviation within different intervals, improving the accuracy of pricing.

[0062] In one possible implementation, the piecewise linear function fitting module includes: a coal profit objective function construction module for constructing a coal profit objective function; an evaluation and screening module for evaluating and screening the first coal pricing dataset and the second coal pricing dataset using the coal profit objective function to obtain a first pricing optimization dataset and a second pricing optimization dataset; and a fitting module for performing piecewise linear function fitting based on the first pricing optimization dataset and the second pricing optimization dataset to determine the dynamic pricing strategy.

[0063] Specifically, coal cost data (such as procurement costs and transportation costs) and historical sales data, including sales prices and sales volumes, are collected. A coal profit objective function is constructed, taking into account the impact of calorific value deviation on the sales price. For example, the coal profit objective function can be expressed as: Profit = (Sales Price - Cost) × Sales Volume, where the sales price is adjusted according to the calorific value deviation.

[0064] The constructed coal profit objective function is used to evaluate the first and second coal pricing datasets, calculate their corresponding profit values, and use optimization algorithms to select the data points that contribute the most to profit, forming the first and second pricing optimization datasets. The first pricing optimization dataset is analyzed to determine the interval division of calorific value deviation, and linear regression is used to fit the relationship between calorific value deviation and price adjustment for each interval. A similar analysis is performed on the second pricing optimization dataset to determine the interval division of calorific value deviation, and linear regression is used to fit the relationship between calorific value deviation and price adjustment for each interval. The piecewise linear functions obtained are integrated into a dynamic pricing strategy. In this implementation, the coal profit objective function construction module, by constructing the profit objective function, can clearly define the profit maximization objective, providing a scientific basis for subsequent data selection and pricing strategies. The evaluation and selection module, by using the profit objective function to evaluate and select historical data, can ensure that the selected data points contribute the most to profit, thereby improving the effectiveness of the pricing strategy.

[0065] In one possible implementation, the fitting module includes: a first fitting module, used to perform piecewise linear function fitting based on the first pricing optimization dataset to obtain a first dynamic function for calorific value pricing; a second fitting module, used to perform piecewise linear function fitting based on the second pricing optimization dataset to obtain a second dynamic function for calorific value pricing; and a result determination module, used to determine the dynamic pricing strategy based on the first dynamic function for calorific value pricing and the second dynamic function for calorific value pricing.

[0066] Specifically, piecewise linear function fitting is performed on the first pricing optimization dataset (calorific value deviation greater than zero). By analyzing the data distribution, the calorific value deviation is divided into multiple intervals, and each interval is fitted with a linear function. The fitted functions of all intervals are then integrated into the first dynamic calorific value pricing function. Piecewise linear function fitting is performed on the second pricing optimization dataset (calorific value deviation less than zero). By analyzing the data distribution, the calorific value deviation is divided into multiple intervals, and each interval is fitted with a linear function. The fitted functions of all intervals are then integrated into the second dynamic calorific value pricing function. The first and second dynamic calorific value pricing functions are integrated into a complete dynamic pricing strategy. The appropriate dynamic pricing function is selected for price adjustment based on the sign of the calorific value deviation.

[0067] This application employs various technical means, including deploying an online LIBS analyzer and a microwave moisture meter on the coal production line to collect production data in real time; using the bomb calorimetry method to obtain calorific value data at preset intervals and mapping it into a set of data pairs; introducing boiler forward and reverse balance data to train and correct the calorific value prediction model to build a prediction channel; and constructing a dynamic pricing strategy based on the real-time calorific value data output from the prediction channel to achieve pricing optimization. These methods solve the technical problem of pricing deviation caused by untimely detection in existing coal pricing systems and achieve the technical effect of dynamic optimization of coal pricing.

[0068] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0069] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A pricing dynamic optimization system based on real-time coal calorific value detection, characterized in that, The system includes: The data acquisition module is used to deploy an online LIBS analyzer and a microwave moisture meter on the coal production line, and to acquire coal production data streams in real time through the online LIBS analyzer and microwave moisture meter. The data association module is used to perform coal calorific value detection using the oxygen bomb calorimetry according to a preset acquisition cycle, obtain coal calorific value data stream, associate and map the coal production data stream and the coal calorific value data stream, and establish a set of coal calorific value data pairs. The calorific value prediction module is used to introduce boiler positive and negative balance data, perform calorific value prediction training and dynamic interference correction on the set based on the coal calorific value data, and build a coal calorific value prediction channel. The pricing optimization module is used to output real-time coal calorific value data through the coal calorific value prediction channel, construct a dynamic pricing strategy, and dynamically optimize the pricing of the real-time coal calorific value data based on the dynamic pricing strategy.

2. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 1, characterized in that, The data association module includes: The coal data distribution threshold determination module is used to determine the coal data distribution threshold according to the coal data application standards. An abnormal data cleaning module is used to clean the coal production data stream and the coal calorific value data stream using the coal data distribution threshold, so as to obtain an available coal production data stream and an available coal calorific value data stream. The time sequence arrangement module is used to arrange the available coal production data stream and the available coal calorific value data stream in time sequence to obtain the coal production sequence data stream and the coal calorific value sequence data stream. The association mapping module is used to perform time alignment and association mapping on the coal production sequence data stream and the coal calorific value sequence data stream, and to establish the set of coal calorific value data pairs.

3. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 1, characterized in that, The calorific value prediction module includes: The set partitioning module is used to proportionally partition the coal calorific value data set to obtain a coal calorific value training set, a coal calorific value verification set, and a coal calorific value test set. The calorific value prediction training module is used to train the coal calorific value prediction on the coal calorific value training set using a random forest network structure, and generate an initial calorific value prediction channel. The iterative parameter tuning module is used to perform verification testing and evaluation of the initial calorific value prediction channel and iterative parameter tuning using the coal calorific value verification set and the coal calorific value test set to obtain the basic calorific value prediction channel. The dynamic interference correction module is used to introduce boiler forward and reverse balance data to dynamically correct the interference of the basic calorific value prediction channel, thereby obtaining the coal calorific value prediction channel.

4. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 3, characterized in that, The dynamic interference correction module includes: The positive balance efficiency calculation module is used to extract input and output heat data and calculate efficiency ratios from the boiler positive and negative balance data to obtain the boiler positive balance efficiency data. The reverse balance efficiency calculation module is used to extract heat loss data and calculate efficiency loss from the boiler forward and reverse balance data to obtain boiler reverse balance efficiency data. The deviation comparison module is used to compare the deviation between the boiler positive balance efficiency data and the boiler negative balance efficiency data to obtain the positive and negative balance efficiency deviation parameters. The channel correction module is used to dynamically correct the basic calorific value prediction channel based on the positive and negative balance efficiency deviation parameters, so as to obtain the coal calorific value prediction channel.

5. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 4, characterized in that, The channel correction module includes: The deviation interference threshold setting module is used to set the deviation interference threshold based on boiler design parameters and actual operating experience. The correction fitting module is used to perform correction fitting based on historical boiler operating data and establish a balance deviation parameter-calorific value correction coefficient function if the positive and negative balance efficiency deviation parameter is greater than the deviation interference threshold. The dynamic calculation module is used to dynamically calculate the positive and negative balance efficiency deviation parameter using the balance deviation parameter-calorific value correction coefficient function, and determine the target calorific value correction coefficient. The interference correction module is used to dynamically correct the interference of the basic calorific value prediction channel based on the target calorific value correction coefficient, so as to obtain the coal calorific value prediction channel.

6. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 5, characterized in that, The dynamic calculation module includes: The absolute value calculation module is used to dynamically calculate the absolute value of the positive and negative balance efficiency deviation parameter using the balance deviation parameter-calorific value correction coefficient function, and obtain the scalar calorific value correction coefficient. The correction direction determination module is used to determine the calorific value correction direction based on the positive and negative balance efficiency deviation parameters. The vector marking module is used to vector mark the scalar calorific value correction coefficient according to the calorific value correction direction, and to determine the target calorific value correction coefficient.

7. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 6, characterized in that, The pricing optimization module includes: The calorific value comparison module is used to obtain the contract benchmark calorific value, compare the predicted calorific value of coal with the contract benchmark calorific value, and obtain the coal calorific value comparison result. The dynamic pricing strategy acquisition module is used to perform historical coal pricing data mining and piecewise linear function fitting based on the coal calorific value comparison results to obtain a dynamic pricing strategy.

8. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 7, characterized in that, The dynamic pricing strategy acquisition module includes: The historical coal pricing data mining module is used to perform historical coal pricing data mining based on the coal calorific value comparison results to obtain a first coal pricing dataset and a second coal pricing dataset. The first coal pricing dataset is a historical pricing dataset in which the real-time calorific value of coal is greater than the contract benchmark calorific value, and the second coal pricing dataset is a historical pricing dataset in which the real-time calorific value of coal is less than the contract benchmark calorific value. The piecewise linear function fitting module is used to perform piecewise linear function fitting using the first coal pricing dataset and the second coal pricing dataset respectively, in order to determine the dynamic pricing strategy.

9. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 8, characterized in that, The piecewise linear function fitting module includes: The coal profit objective function construction module is used to construct the coal profit objective function; The evaluation and screening module is used to evaluate and screen the first coal pricing dataset and the second coal pricing dataset using the coal profit objective function to obtain the first pricing optimization dataset and the second pricing optimization dataset. The fitting module is used to perform piecewise linear function fitting based on the first pricing optimization dataset and the second pricing optimization dataset to determine the dynamic pricing strategy.

10. The pricing dynamic optimization system based on real-time coal calorific value detection as described in claim 9, characterized in that, The fitting module includes: The first fitting module is used to perform piecewise linear function fitting based on the first pricing optimization dataset to obtain the first dynamic function for calorific value pricing. The second fitting module is used to perform piecewise linear function fitting based on the second pricing optimization dataset to obtain the second calorific value pricing dynamic function. The result determination module is used to determine the dynamic pricing strategy based on the first calorific value pricing dynamic function and the second calorific value pricing dynamic function.

Images