Coking product yield correction method, system, and electronic device

Abstract

The present application relates to the technical field of coking, and particularly relates to a coking product yield correction method and system and an electronic device, the method comprising: obtaining historical production data of a coking product, the historical production data comprising production data of consecutive days, the production data comprising production condition data and product yield observation values; performing scene division on all production data based on the production condition data and a preset scene classification rule, to determine a production scene to which daily production data belongs; traversing the product yield observation values under each production scene, to construct a probability density prediction model, the probability density prediction model corresponding to the production scene in a one-to-one manner; constructing an optimization model according to the probability density prediction model, and using the optimization model to complete coking product yield correction; the method can accurately correct the coking product yield, and the method helps to improve the prediction accuracy of the coking product yield under different production scenes.

Description

Technical Field

[0001] This invention relates to the field of coking technology, and in particular to a method, system and electronic equipment for correcting the yield of coking products. Background Technology

[0002] Coking is the process of pyrolyzing or cracking organic materials (usually coal or petroleum) under high-temperature conditions. Coking produces gaseous, liquid, and solid products (such as coke). To ensure the continuity of the entire coking process and the automated control of production equipment, accurate and timely statistics on coking product output are essential.

[0003] Currently, manual recording is commonly used to statistically record the output of coking products. However, this method inevitably leads to recording errors, such as data transcription mistakes, which can cause data anomalies during coking production and even production delays. Summary of the Invention

[0004] This invention provides a method, system, and electronic device for correcting the output of coking products, in order to solve the problem that the existing technology uses manual recording to statistically record the output of coking products, which inevitably leads to recording errors, such as data copying errors, resulting in data anomalies during the coking production process, and even causing production delays.

[0005] This invention provides a method for correcting the yield of coking products, the method comprising:

[0006] Acquire historical production data of coking products, including production data for multiple consecutive days, and the production data includes: production condition data and product output observations;

[0007] Based on the production condition data and the preset scenario classification rules, all the production data are divided into scenarios to determine the production scenario to which the production data belongs each day.

[0008] The product output observations under each production scenario are traversed to construct a probability density prediction model. The probability density prediction model corresponds one-to-one with each production scenario and describes the probability distribution of the output of coking products under the corresponding production scenario.

[0009] An optimization model is constructed based on the probability density prediction model, and the optimization model is used to correct the output of coking products.

[0010] In one embodiment of the present invention, the step of setting the scene classification rules includes:

[0011] Based on preset production condition indicators, target data corresponding to the production condition indicators are selected from all the production condition data. There is at least one production condition indicator, and all the target data corresponding to one production condition indicator form a target data group.

[0012] By dividing all the target data groups into intervals, at least one interval segment is obtained for each target data group;

[0013] Each target data group selects one interval segment for combination to obtain the production scenario. The production scenario includes a scenario number and an interval segment, and the interval segments in the production scenario correspond one-to-one with the production condition indicators.

[0014] Based on all the obtained production scenarios, the scenario classification rules are determined.

[0015] In one embodiment of the present invention, the step of traversing the product output observations under each production scenario to construct a probability density prediction model includes:

[0016] Obtain the mean and standard deviation of the product output observations under any of the aforementioned production scenarios;

[0017] Based on the mean, the standard deviation, and the preset first probability distribution function, a probability density prediction model for the current production scenario is constructed; or, based on the mean and the preset second probability distribution function, a probability density prediction model for the current production scenario is constructed.

[0018] Given the probability density prediction model for the current production scenario, the product output observations for the remaining production scenarios are iterated through to obtain the probability density prediction models for the remaining production scenarios.

[0019] In one embodiment of the present invention, the first probability distribution function is a Gaussian distribution function; based on the mean, the standard deviation, and the first probability distribution function, the mathematical expression for constructing the probability density prediction model under the current production scenario is as follows:

[0020]

[0021] Where λ represents the mean of the observed product output in the i-th production scenario, and σ i Let f(x) represent the standard deviation of the observed product output in the i-th production scenario, e represent the natural constant, and f(x) represent the standard deviation of the observed product output. i () indicates that, based on a probability density prediction model constructed using a Gaussian distribution function, the output of coking products in the i-th production scenario is x. iThe probability; the current mathematical expression represents the probability distribution of the output of coking products in the i-th production scenario, which follows a mean of λ and a standard deviation of σ. i The Gaussian distribution.

[0022] In one embodiment of the present invention, the second probability distribution function is an exponential distribution function; based on the mean and the preset second probability distribution function, the mathematical expression for constructing the probability density prediction model under the current production scenario is as follows:

[0023]

[0024] Where λ represents the mean of the observed product output in the i-th production scenario, e represents the natural constant, and f(x) i )′ represents the output of coking products in the i-th production scenario as x, based on a probability density prediction model constructed using the exponential distribution function. i The probability; the current mathematical expression indicates that, in the i-th production scenario, the probability distribution of the output of coking products follows an exponential distribution with a mean of λ.

[0025] In one embodiment of the present invention, the optimization model includes: an objective function, a first constraint, and a second constraint; the objective function is used to obtain the objective solution of the decision variables of the probability density prediction model, wherein the decision variables refer to the correction value variables of the output of coking products, and the objective solution is obtained by determining the value of the corresponding decision variable as the objective solution of the decision variable when the probability product of all decision variables is maximized; the first constraint is used to restrict the value of a single decision variable to a first interval, wherein the first interval is determined based on the mean and standard deviation of the product output observations under the current production scenario; the second constraint is used to restrict the sum of the values ​​of all decision variables to a second interval, wherein the second interval is determined based on the sum of the product output observations over multiple consecutive days;

[0026] The steps for using the optimization model to correct coking product yield include:

[0027] Based on the objective function, the first constraint, and the second constraint, a corrected value for the output of coking products over several consecutive days is obtained, and the corrected value is the output value of coking products after correction.

[0028] In one embodiment of the present invention, the step of determining the first interval includes:

[0029] The first intermediate value is determined by multiplying the standard deviation of the observed product output in the current production scenario with the preset target weight.

[0030] The average value of the observed product output under the current production scenario is determined as the second median value;

[0031] The sum of the first intermediate value and the second intermediate value is determined as the maximum value of the first interval;

[0032] The difference between the second intermediate value and the first intermediate value is determined as the minimum value of the first interval.

[0033] In one embodiment of the present invention, the step of determining the second interval includes:

[0034] The sum of all product output observations is determined as the third intermediate value; the production scenario to which the production data of the first day in the historical production data belongs is determined as the first production scenario, the standard deviation of the product output observations of the first production scenario is determined as the first standard deviation, the production scenario to which the production data of the last day in the historical production data belongs is determined as the second production scenario, the standard deviation of the product output observations of the second production scenario is determined as the second standard deviation; the sum of the first standard deviation and the second standard deviation is determined as the fourth intermediate value; the product between the fourth intermediate value and the target weight is determined as the fifth intermediate value, and the difference between the third intermediate value and the fifth intermediate value is determined as the minimum value of the second interval;

[0035] The production scenario to which the production data belongs on the day preceding the historical production data is defined as the third production scenario, and the standard deviation of the product output observations in the third production scenario is defined as the third standard deviation. The production scenario to which the production data belongs on the day following the historical production data is defined as the fourth production scenario, and the standard deviation of the product output observations in the fourth production scenario is defined as the fourth standard deviation. The sum of the third and fourth standard deviations is defined as the sixth median. The product of the sixth median and the target weight is defined as the seventh median. The sum of the seventh median and the third median is defined as the maximum value of the second interval.

[0036] The present invention also provides a coking product yield correction system, comprising:

[0037] The data acquisition module is used to acquire historical production data of coking products. The historical production data includes production data for multiple consecutive days, and the production data includes: production condition data and product output observations.

[0038] The scenario segmentation module is used to segment all the production data into scenarios based on the production condition data and preset scenario classification rules, so as to determine the production scenario to which the production data of each day belongs.

[0039] The prediction model building module is used to traverse the product output observations under each production scenario to build a probability density prediction model. The probability density prediction model corresponds one-to-one with the production scenario and describes the probability distribution of the output of coking products under the corresponding production scenario.

[0040] The correction module is used to construct an optimization model based on the probability density prediction model, and to use the optimization model to complete the correction of coking product output.

[0041] The present invention also provides an electronic device, including a processor, a memory, and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement the coking product yield correction method provided in any of the above embodiments.

[0042] The beneficial effects of this invention are as follows: The coking product output correction method, system, and electronic equipment proposed in this invention acquire historical production data of coking products, including production data from multiple consecutive days. This production data includes production condition data and observed product output values. Based on the production condition data and preset scenario classification rules, all production data is divided into scenarios to determine the production scenario to which the daily production data belongs. The observed product output values ​​under each production scenario are traversed to construct a probability density prediction model. This probability density prediction model corresponds one-to-one with the production scenario and describes the probability distribution of coking product output under the corresponding production scenario. An optimization model is constructed based on the probability density prediction model, and the optimized model is used to complete the coking product output correction. This method can accurately correct the coking product output. Furthermore, this method takes into account different production scenarios in coking production, which helps improve the prediction accuracy of coking product output under different production scenarios, thereby improving the accuracy of coking product output correction. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating a method for correcting the yield of coking products according to an embodiment of the present invention.

[0044] Figure 2 This is a schematic diagram of the structure of a coking product yield correction system provided in one embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0046] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0047] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0048] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0049] The following is combined Figures 1 to 3 The present invention provides an explanation of the coking product yield correction method, system, and electronic equipment provided by the present invention.

[0050] Please see Figure 1 , Figure 1 This is a flowchart illustrating a method for correcting coking product yield according to an embodiment of the present invention, as shown below. Figure 1 As shown, the coking product yield correction method includes:

[0051] S110: Obtain historical production data of coking products, the historical production data including production data for multiple consecutive days, the production data including: production condition data and product output observations.

[0052] It should be noted that the production condition data refers to production data generated during the coking process, such as production process parameters and quality inspection indicators. This production condition data can be obtained from historical daily production data. Production process parameters include coking time and execution temperature, while quality inspection indicators include raw coal quality inspection indicators such as volatile matter (Vdaf), sulfur (Sta), and ash (Ash). This embodiment, by acquiring the aforementioned production condition data, facilitates subsequent production scenario segmentation based on this data, assigning daily production data to corresponding production scenarios. Furthermore, by acquiring continuous product output observations over multiple days, this embodiment facilitates the construction of corresponding probability density prediction models based on product output observations under each production scenario, and also facilitates subsequent coking product output correction.

[0053] S120: Based on the production condition data and preset scenario classification rules, all the production data are divided into scenarios to determine the production scenario to which the daily production data belongs.

[0054] Specifically, by matching production condition data with production scenarios in the scenario classification rules, the production scenario to which each day's production data belongs is determined. It should be noted that by dividing all production data into scenarios, each production scenario corresponds to at least one day of production data.

[0055] S130: Traverse the product output observations under each production scenario to construct a probability density prediction model. The probability density prediction model corresponds one-to-one with each production scenario and describes the probability distribution of coking product output under the corresponding production scenario.

[0056] It should be noted that, for any production scenario, by traversing all product output observations under that production scenario, the distribution of product output observations under that production scenario can be obtained. Then, based on the distribution of product output observations under that production scenario, an appropriate probability distribution (such as Gaussian distribution and exponential distribution) can be selected to complete the construction of the probability density prediction model.

[0057] It should also be noted that by constructing probability density prediction models that correspond one-to-one with production scenarios, we can better adapt to various uncertainties and complexities in the coking product production process, effectively ensure the accuracy of coking product output correction for different production scenarios, and have a high degree of flexibility.

[0058] It's worth mentioning that the probability density prediction model can effectively describe the probability distribution of coking product output under corresponding production scenarios. Compared to directly predicting a single correction value, this probability distribution better reflects the uncertainty and volatility of coking product output. Therefore, whether under normal fluctuations or special circumstances, this probability distribution can more accurately describe the possible range of coking product output, reducing the risks associated with data correction based on a single correction value, minimizing the possibility of misjudgments and omissions, and improving the accuracy of coking product output correction.

[0059] S140: Construct an optimization model based on the probability density prediction model, and use the optimization model to complete the correction of coking product output.

[0060] It should be noted that the optimization model is a mathematical formalization aimed at solving optimization problems, namely, finding variable values ​​that maximize or minimize the objective function under given constraints. This embodiment constructs an optimization model based on a probability density prediction model and uses this model to correct the output of coking products. This achieves better target optimization, obtaining a better corrected value for the output of coking products under given constraints, thus realizing the correction of coking product output.

[0061] It should also be noted that coking production, such as delayed coking processes, is highly complex, and its product output is highly volatile. Many variables in the coking process (such as raw material properties, temperature, and pressure) can affect the yield of coking products. Therefore, the above embodiments, through steps such as dividing production scenarios (each corresponding to different production condition data, i.e., any value or range of the aforementioned variables), constructing probability density prediction models, and building optimization models, can better adapt to the randomness and uncertainty in the coking production process.

[0062] In some embodiments, the steps for setting the scene classification rules include:

[0063] 1. Based on preset production condition indicators, target data corresponding to the production condition indicators are selected from all the production condition data. There is at least one production condition indicator, and all the target data corresponding to one production condition indicator form a target data group.

[0064] It should be noted that the production condition indicators can be set according to actual conditions, such as coking time, operating temperature, volatile matter, sulfur content, production plan, and equipment maintenance plan. During the specific setting process, appropriate production condition indicators can be selected from the coking product production data as dimensions for production scenario classification, based on actual conditions.

[0065] Second, by dividing all the target data groups into intervals, at least one interval segment is obtained for each target data group. For example, for target data group one, the maximum and minimum values ​​in target data group one constitute an interval. By dividing this interval, at least one interval segment of target data group one is obtained. The number of interval segments can be set according to actual needs, such as 2 segments or 3 segments.

[0066] Third, select one interval from each of the target data groups and combine them to obtain the production scenario. The production scenario includes a scenario number and an interval, and the interval in the production scenario corresponds one-to-one with the production condition indicators.

[0067] Specifically, select an interval from each target data group, combine the selected intervals to form a production scenario, and each production scenario has a corresponding scenario number.

[0068] Fourth, based on all the obtained production scenarios, determine the scenario classification rules. Specifically, all the obtained production scenarios are determined as scenario classification rules. By obtaining the scenario classification rules, it is possible to achieve a better and more refined division of production scenarios.

[0069] For example: assuming hs i Let represent the i-th production scenario, and let HS and hs be the set of production scenarios. i ∈HS, HS={hs i Let , i∈N}, where N represents the number of production scenarios. Below is an example of a production scenario: Production Scenario 1: Simultaneously satisfying 24h (hours) < coking time ≤ 25h, 1400 < execution temperature ≤ 1405, 22 < volatile matter ≤ 23, 0.5 < sulfur content ≤ 1, its mathematical expression is: hs1={24 <cokingTime≤25,1400<execuTem≤1405,22<Vdaf≤23,0.5<Sta≤1}。

[0070] In some embodiments, the step of iterating through the product output observations for each production scenario to construct a probability density prediction model includes:

[0071] 1. Obtain the mean and standard deviation of the product output observations under any of the aforementioned production scenarios.

[0072] Specifically, the mathematical expression for obtaining the mean is as follows:

[0073]

[0074] Where λ represents the mean of the observed product output in the i-th production scenario. Let represent the output observation of the j-th product in the i-th production scenario, and n represent the number of output observations in the i-th production scenario.

[0075] The mathematical expression for obtaining the standard deviation is:

[0076]

[0077] Where, σ i Let represent the standard deviation of the observed product output in the i-th production scenario.

[0078] 2. Based on the mean, the standard deviation, and the preset first probability distribution function, construct a probability density prediction model for the current production scenario; or, based on the mean and the preset second probability distribution function, construct a probability density prediction model for the current production scenario.

[0079] It should be noted that constructing a probability density prediction model for the current production scenario based on the mean, standard deviation, and a first probability distribution function, or based on the mean and a second probability distribution function, can yield a highly accurate probability density prediction model. Here, the first probability distribution function is a function that distributes the probability of coking product output based on the mean and standard deviation, while the second probability distribution function is a function that distributes the probability of coking product output based on the mean.

[0080] Third, given the probability density prediction model for the current production scenario, iterate through the product output observations for the remaining production scenarios to obtain the probability density prediction model for the remaining production scenarios.

[0081] To further improve the accuracy of the probability density prediction model, in some embodiments, the first probability distribution function is a Gaussian distribution function; based on the mean, the standard deviation, and the first probability distribution function, the mathematical expression of the probability density prediction model for the current production scenario is as follows:

[0082]

[0083] Where λ represents the mean of the observed product output in the i-th production scenario, and σ i Let f(x) represent the standard deviation of the observed product output in the i-th production scenario, e represent the natural constant, and f(x) represent the standard deviation of the observed product output. i () indicates that, based on a probability density prediction model constructed using a Gaussian distribution function, the output of coking products in the i-th production scenario is x. i The probability; the current mathematical expression represents the probability distribution of the output of coking products in the i-th production scenario, which follows a mean of λ and a standard deviation of σ. iThe model uses a Gaussian distribution. It should be noted that by employing a Gaussian distribution function, a probability density prediction model can be constructed for the current production scenario, satisfying various production scales and product output distributions. This model has strong universality and high accuracy.

[0084] In some embodiments, the second probability distribution function is an exponential distribution function; based on the mean and the preset second probability distribution function, the mathematical expression for constructing the probability density prediction model under the current production scenario is as follows:

[0085]

[0086] Where λ represents the mean of the observed product output in the i-th production scenario, e represents the natural constant, and f(x) i )′ represents the output of coking products in the i-th production scenario as x, based on a probability density prediction model constructed using the exponential distribution function. i The probability is expressed mathematically as follows: in the i-th production scenario, the probability distribution of coking product output follows an exponential distribution with mean λ. It should be noted that constructing a probability density prediction model for the current production scenario based on the exponential distribution function can significantly improve the accuracy of the probability density prediction model.

[0087] The above embodiments completed the construction of probability density prediction models for each production scenario. The construction steps of the optimization model will be explained below.

[0088] In some embodiments, the optimization model includes: an objective function, a first constraint, and a second constraint; the objective function is used to obtain the objective solution of the decision variables of the probability density prediction model, the decision variables refer to the correction value variables of the output of coking products, and the objective solution is obtained by determining the value of the corresponding decision variable as the objective solution of the decision variable when the probability product of all decision variables is maximized; the first constraint is used to restrict the value of a single decision variable to a first interval, the first interval being determined based on the mean and standard deviation of the product output observations under the current production scenario; the second constraint is used to restrict the sum of the values ​​of all decision variables to a second interval, the second interval being determined based on the sum of the product output observations over multiple consecutive days.

[0089] It should be noted that constructing the aforementioned optimization model helps to optimize the decision variables of the probability density prediction model. Specifically, the objective function in this embodiment is obtained based on the probability density prediction model, which is equivalent to a data model. Therefore, based on the objective function, the objective solution for the decision variables can be obtained. By setting a first constraint, a single objective solution can be restricted to a first interval to ensure the rationality of a single objective solution. By setting a second constraint, the sum of all objective solutions, i.e., the sum of the values ​​of all decision variables, can be restricted to a second interval, thereby ensuring the stability of the predicted total output.

[0090] Furthermore, the steps for using the optimization model to correct the coking product yield include:

[0091] Based on the objective function, the first constraint, and the second constraint, a corrected value for the output of coking products over several consecutive days is obtained, and the corrected value is the output value of coking products after correction.

[0092] Specifically, the mathematical expression of the objective function is as follows:

[0093]

[0094] Where max represents finding the maximum value, and T represents the number of product output observations, i.e. the number of days of production data.

[0095] In some embodiments, the step of determining the first interval includes:

[0096] First, the product of the standard deviation of the observed product output in the current production scenario and the preset target weight is determined as the first intermediate value.

[0097] Secondly, the mean of the observed product output under the current production scenario is determined as the second median value.

[0098] Then, the sum of the first intermediate value and the second intermediate value is determined as the maximum value of the first interval.

[0099] Finally, the difference between the second intermediate value and the first intermediate value is determined as the minimum value of the first interval.

[0100] Specifically, the mathematical expression of the first constraint is as follows:

[0101] λ-K×σ i ≤x i ≤λ+K×σ i

[0102] Where K represents the target weight, the value of K can be set according to actual needs, such as [1,3] etc.

[0103] In some embodiments, the step of determining the second interval includes:

[0104] 1. The sum of all product output observations is determined as the third intermediate value; the production scenario to which the production data of the first day in the historical production data belongs is determined as the first production scenario, the standard deviation of the product output observations of the first production scenario is determined as the first standard deviation, the production scenario to which the production data of the last day in the historical production data belongs is determined as the second production scenario, the standard deviation of the product output observations of the second production scenario is determined as the second standard deviation; the sum of the first standard deviation and the second standard deviation is determined as the fourth intermediate value; the product between the fourth intermediate value and the target weight is determined as the fifth intermediate value, and the difference between the third intermediate value and the fifth intermediate value is determined as the minimum value of the second interval.

[0105] Second, the production scenario to which the production data of the day preceding the historical production data belongs is determined as the third production scenario; the standard deviation of the product output observation value of the third production scenario is determined as the third standard deviation; the production scenario to which the production data of the day following the historical production data belongs is determined as the fourth production scenario; the standard deviation of the product output observation value of the fourth production scenario is determined as the fourth standard deviation; the sum of the third standard deviation and the fourth standard deviation is determined as the sixth median; the product between the sixth median and the target weight is determined as the seventh median; the sum of the seventh median and the third median is determined as the maximum value of the second interval.

[0106] Specifically, the mathematical expression of the second constraint is as follows:

[0107]

[0108] Among them, w i Let σi represent the output observation of the i-th product, σ1 represent the standard deviation of the output observations of the first production scenario, and σi represent the standard deviation of the output observations of the product. T σ0 represents the standard deviation of the product output observations in the second production scenario, and σ0 represents the standard deviation of the product output observations in the third production scenario, which is determined to be the third standard deviation. T+1 This represents the standard deviation of the observed product output in the fourth production scenario.

[0109] It should be noted that the production data for the day preceding the historical production data is known data, and the production data for the day following the historical production data is either known or unknown data. If the production data for the day following the historical production data is unknown, then σ T+1 The value of is 0.

[0110] It is also worth mentioning that by setting the first and second intervals mentioned above, the correlation between data can be well considered while optimizing the daily coking product output data, that is, the cumulative effect of data over multiple days. This ensures that the corrected coking product output data is accurate on a single day, and the accumulated value can also accurately reflect the actual total production.

[0111] The optimization model, in correcting coking product output, considers not only the rationality of single-day data but also treats data from multiple consecutive days as a whole for optimization. This ensures that while maintaining the rationality of each individual daily data point, the sum of the corrected daily data still aligns with the total output observed over the continuous time period. This approach effectively prevents data conflicts between different dates caused by single-day data correction, enhancing the integrity and consistency of the data correction and thus more accurately reflecting the actual operating status of the coking production process.

[0112] It should also be mentioned that the prediction and optimization in the above embodiments apply to all production data. In actual implementation, production data from multiple consecutive days can also be selected from all production data as the object of prediction and optimization. The prediction and optimization methods are the same as above, and will not be repeated here.

[0113] Compared to rule-based verification schemes (such as setting a "normal fluctuation range" and judging whether the observed product output falls within this range to correct it), the coking product output correction method in the above embodiments can complete the output correction without defining a "normal fluctuation range," effectively reducing errors caused by human intervention and achieving higher accuracy. It is understandable that due to the significant volatility and instability in the coking product production process, it is difficult to accurately define the "normal fluctuation range" of coking product output, and the rule-based verification schemes described above are prone to misjudgment.

[0114] It's worth mentioning that the rule-based verification scheme described above might affect the continuity of data between two adjacent days when correcting data anomalies on a single day. For example, if the product output observation from the previous day is mistakenly included in the current day's data, using the rule-based method to correct both the current and previous day's output might result in the corrected output for the current day being lower than the actual value, while the corrected output for the previous day might be higher. This would lead to a lack of smooth transition between the corrected product output observations for two consecutive days, failing to accurately reflect the continuous changes in the actual production process. The coking product output correction method in the above embodiments, by constructing an optimized model, effectively solves this problem. It ensures the accuracy of single-day data correction while also considering the continuity between data from multiple days, guaranteeing that the sum of the corrected output values ​​for a single day accurately reflects the actual total production.

[0115] Furthermore, compared to correction schemes based on statistical models and time series analysis (establishing statistical models based on historical data, such as ARIMA (Autoregressive Integrated Moving Average Model) models, state-space models, etc.; making time series predictions of product output; and identifying and correcting abnormal data included across periods by comparing model predictions with actual observations), the coking product output correction method in the above embodiments can better consider special circumstances in the coking product production process, thereby enhancing the accuracy of data correction. It is understandable that correction schemes based on statistical models and time series analysis often predict future data based on historical data. However, their prediction results often ignore special or unexpected situations in the actual production process, such as equipment failures or changes in raw material properties. When such special circumstances occur, this scheme is very likely to produce prediction anomalies. The coking product output correction method in the above embodiments, by dividing multi-day production data into scenarios and establishing corresponding probability density prediction models for each production scenario, can better overcome the impact of the aforementioned special circumstances on model predictions, demonstrating strong adaptability.

[0116] Furthermore, the aforementioned correction schemes based on statistical models and time series analysis also suffer from the problem of data mismatch. Specifically, when correcting inter-period data, the model-predicted total output may differ from the actual total output. However, the coking product output correction method in the above embodiments, by constructing an optimized model, effectively overcomes this problem. It can correct single-day data while ensuring the continuity and consistency of multi-day data, effectively mitigating the problems caused by data delays (such as including the previous day's product output observation in today's data) and inter-period inclusion.

[0117] In summary, the coking product output correction method in the above embodiments, by introducing a probability density prediction model (which corresponds one-to-one with the production scenario) and an optimization model, not only improves the accuracy and stability of coking product output data correction, but also enhances its adaptability to complex industrial production environments and the consistency of data correction, making it highly feasible.

[0118] The following specific embodiment will be used to explain the coking product yield correction method in the above embodiment.

[0119] First, obtain historical production data for coking products. Historical production data includes production data from multiple consecutive days, and includes production condition data and observed product output values.

[0120] Secondly, determine the scene classification rules, that is, to subdivide the production scene, such as based on four dimensions: coking time, execution temperature, volatile matter and sulfur content, to obtain multiple production scenes.

[0121] Then, the production data is divided into scenarios to determine the production scenario described in the daily production data.

[0122] Next, a probability density prediction model corresponding to each production scenario is established.

[0123] Next, the optimization model is established. The methods for constructing the probability density prediction model and the optimization model are described above and will not be repeated here. It is worth mentioning that in the actual optimization process, production data for a continuous target number of days can also be selected from all production data for model construction. Correspondingly, subsequent optimization and correction will only be performed on the production data for that selected target number of days.

[0124] Finally, the optimization model is solved using a genetic algorithm and a particle swarm optimization algorithm to obtain the corrected values ​​of the output of coking products over several consecutive days. These corrected values ​​are the output values ​​of the coking products after correction.

[0125] The coking product yield correction system provided by the present invention is described below. The coking product yield correction system described below can be referred to in correspondence with the coking product yield correction method described above.

[0126] Please refer to Figure 2 The coking product yield correction system provided in this embodiment includes:

[0127] Data acquisition module 210 is used to acquire historical production data of coking products. The historical production data includes production data for multiple consecutive days and includes production condition data and product output observations.

[0128] The scenario segmentation module 220 is used to segment all the production data into scenarios based on the production condition data and preset scenario classification rules, so as to determine the production scenario to which the production data of each day belongs.

[0129] The prediction model building module 230 is used to traverse the product output observations under each production scenario to build a probability density prediction model. The probability density prediction model corresponds one-to-one with the production scenario and describes the probability distribution of the output of coking products under the corresponding production scenario.

[0130] The correction module 240 is used to construct an optimized model based on the probability density prediction model, and to use the optimized model to complete the correction of coking product output. The data acquisition module 210, the scene segmentation module 220, the prediction model construction module 230, and the correction module 240 are connected. The coking product output correction system in this embodiment can accurately correct the coking product output. Furthermore, this method takes into account different production scenarios in coking production, which helps to improve the prediction accuracy of coking product output under different production scenarios, thereby improving the accuracy of coking product output correction, and has high feasibility.

[0131] In some embodiments, it further includes: a scenario classification rule setting module, used to filter out target data corresponding to the production condition indicators from all the production condition data based on preset production condition indicators, wherein the production condition indicators are at least one, and all the target data corresponding to a production condition indicator constitute a target data group.

[0132] By dividing all the target data groups into intervals, at least one interval segment is obtained for each target data group;

[0133] Each target data group selects one interval segment for combination to obtain the production scenario. The production scenario includes a scenario number and an interval segment, and the interval segments in the production scenario correspond one-to-one with the production condition indicators.

[0134] Based on all the obtained production scenarios, the scenario classification rules are determined.

[0135] In some embodiments, the prediction model building module 230 is specifically used to obtain the mean and standard deviation of the product output observations under any of the production scenarios;

[0136] Based on the mean, the standard deviation, and the preset first probability distribution function, a probability density prediction model for the current production scenario is constructed; or, based on the mean and the preset second probability distribution function, a probability density prediction model for the current production scenario is constructed.

[0137] Given the probability density prediction model for the current production scenario, the product output observations for the remaining production scenarios are iterated through to obtain the probability density prediction models for the remaining production scenarios.

[0138] In some embodiments, the correction module 240 is specifically used to obtain a correction value for the output of coking products over several consecutive days based on the objective function, the first constraint, and the second constraint, wherein the correction value is the corrected output value of coking products.

[0139] In some embodiments, the correction module 240 is further configured to determine the product between the standard deviation of the product output observation value in the current production scenario and the preset target weight as the first intermediate value;

[0140] The average value of the observed product output under the current production scenario is determined as the second median value;

[0141] The sum of the first intermediate value and the second intermediate value is determined as the maximum value of the first interval;

[0142] The difference between the second intermediate value and the first intermediate value is determined as the minimum value of the first interval.

[0143] In some embodiments, the correction module 240 is further configured to: determine the sum of all product output observations as a third intermediate value; determine the production scenario to which the production data of the first day in the historical production data belongs as a first production scenario; determine the standard deviation of the product output observations of the first production scenario as a first standard deviation; determine the production scenario to which the production data of the last day in the historical production data belongs as a second production scenario; determine the standard deviation of the product output observations of the second production scenario as a second standard deviation; determine the sum of the first standard deviation and the second standard deviation as a fourth intermediate value; determine the product between the fourth intermediate value and the target weight as a fifth intermediate value; and determine the difference between the third intermediate value and the fifth intermediate value as the minimum value of the second interval.

[0144] The production scenario to which the production data belongs on the day preceding the historical production data is defined as the third production scenario, and the standard deviation of the product output observations in the third production scenario is defined as the third standard deviation. The production scenario to which the production data belongs on the day following the historical production data is defined as the fourth production scenario, and the standard deviation of the product output observations in the fourth production scenario is defined as the fourth standard deviation. The sum of the third and fourth standard deviations is defined as the sixth median. The product of the sixth median and the target weight is defined as the seventh median. The sum of the seventh median and the third median is defined as the maximum value of the second interval.

[0145] In some embodiments, an electronic device is also provided, which may be a server, and its internal structure diagram is shown below. Figure 3As shown, the electronic device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of the server-side method described above.

[0146] In some embodiments, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the following steps: acquiring historical production data of coking products, the historical production data including production data for multiple consecutive days, the production data including: production condition data and product output observations; classifying all production data into scenarios based on the production condition data and preset scenario classification rules to determine the production scenario to which the daily production data belongs; traversing the product output observations under each production scenario to construct a probability density prediction model, the probability density prediction model corresponding one-to-one with the production scenario, the probability density prediction model describing the probability distribution of coking product output under the corresponding production scenario; constructing an optimization model based on the probability density prediction model, and using the optimization model to complete the coking product output correction.

[0147] In some embodiments, a computer-readable storage medium is provided having a computer program stored thereon. When executed by a processor, the computer program performs the following steps: acquiring historical production data of coking products, the historical production data including production data for multiple consecutive days, the production data including: production condition data and product output observations; classifying all production data into scenarios based on the production condition data and preset scenario classification rules to determine the production scenario to which the daily production data belongs; traversing the product output observations under each production scenario to construct a probability density prediction model, the probability density prediction model corresponding one-to-one with the production scenario, the probability density prediction model describing the probability distribution of coking product output under the corresponding production scenario; constructing an optimization model based on the probability density prediction model, and using the optimization model to complete the coking product output correction.

[0148] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or electronic device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0149] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0150] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for correcting the yield of coking products, characterized in that, include: Acquire historical production data of coking products, including production data for multiple consecutive days, and the production data includes: production condition data and product output observations; Based on the production condition data and the preset scenario classification rules, all the production data are divided into scenarios to determine the production scenario to which the production data belongs each day. The product output observations under each production scenario are traversed to construct a probability density prediction model. The probability density prediction model corresponds one-to-one with each production scenario and describes the probability distribution of the output of coking products under the corresponding production scenario. An optimization model is constructed based on the probability density prediction model, and the optimization model is used to complete the coking product output correction. The steps for setting the scene classification rules include: Based on preset production condition indicators, target data corresponding to the production condition indicators are selected from all the production condition data. There is at least one production condition indicator, and all the target data corresponding to one production condition indicator form a target data group. By dividing all the target data groups into intervals, at least one interval segment is obtained for each target data group; Each target data group selects one interval segment for combination to obtain the production scenario. The production scenario includes a scenario number and an interval segment, and the interval segments in the production scenario correspond one-to-one with the production condition indicators. Based on all the obtained production scenarios, the scenario classification rules are determined; The optimization model includes: an objective function, a first constraint, and a second constraint. The objective function is used to obtain the target solution of the decision variables of the probability density prediction model. The decision variables refer to the correction value variables of the output of coking products. The target solution is obtained by determining the value of the corresponding decision variable as the target solution when the product of probabilities of all decision variables is maximized. The first constraint is used to restrict the value of a single decision variable to a first interval, which is determined based on the mean and standard deviation of the product output observations under the current production scenario. The second constraint is used to restrict the sum of the values ​​of all decision variables to a second interval, which is determined based on the sum of the product output observations over multiple consecutive days. The steps for using the optimization model to correct the output of coking products include: obtaining corrected values ​​of the output of coking products over several consecutive days based on the objective function, the first constraint, and the second constraint, wherein the corrected values ​​are the corrected output values ​​of coking products.

2. The method for correcting coking product yield according to claim 1, characterized in that, The steps of iterating through the product output observations for each of the aforementioned production scenarios to construct a probability density prediction model include: Obtain the mean and standard deviation of the product output observations under any of the aforementioned production scenarios; Based on the mean, the standard deviation, and the preset first probability distribution function, a probability density prediction model for the current production scenario is constructed; or, based on the mean and the preset second probability distribution function, a probability density prediction model for the current production scenario is constructed. Given the probability density prediction model for the current production scenario, the product output observations for the remaining production scenarios are iterated through to obtain the probability density prediction models for the remaining production scenarios.

3. The method for correcting coking product yield according to claim 2, characterized in that, The first probability distribution function is a Gaussian distribution function; based on the mean, the standard deviation, and the first probability distribution function, the mathematical expression for constructing the probability density prediction model in the current production scenario is as follows: in, This represents the mean of the observed product output values ​​in the i-th production scenario. Let represent the standard deviation of the observed product output in the i-th production scenario, and let e represent the natural constant. This indicates that, based on a probability density prediction model constructed using a Gaussian distribution function, the output of coking products in the i-th production scenario is... The probability; the current mathematical expression represents that, in the i-th production scenario, the probability distribution of the output of coking products follows a mean of . Standard deviation is The Gaussian distribution.

4. The method for correcting coking product yield according to claim 2, characterized in that, The second probability distribution function is an exponential distribution function; based on the mean and the preset second probability distribution function, the mathematical expression of the probability density prediction model for the current production scenario is as follows: in, Let represent the mean of the observed product output in the i-th production scenario, and let e represent the natural constant. This indicates that, based on a probability density prediction model constructed using the exponential distribution function, the output of coking products in the i-th production scenario is... The probability; the current mathematical expression represents that, in the i-th production scenario, the probability distribution of the output of coking products follows a mean of . The exponential distribution.

5. The method for correcting coking product yield according to claim 1, characterized in that, The steps for determining the first interval include: The first intermediate value is determined by multiplying the standard deviation of the observed product output in the current production scenario with the preset target weight. The average value of the observed product output under the current production scenario is determined as the second median value; The sum of the first intermediate value and the second intermediate value is determined as the maximum value of the first interval; The difference between the second intermediate value and the first intermediate value is determined as the minimum value of the first interval.

6. The method for correcting coking product yield according to claim 1 or 5, characterized in that, The steps for determining the second interval include: The sum of all product output observations is determined as the third intermediate value; the production scenario to which the production data of the first day in the historical production data belongs is determined as the first production scenario, the standard deviation of the product output observations of the first production scenario is determined as the first standard deviation, the production scenario to which the production data of the last day in the historical production data belongs is determined as the second production scenario, the standard deviation of the product output observations of the second production scenario is determined as the second standard deviation; the sum of the first standard deviation and the second standard deviation is determined as the fourth intermediate value; the product between the fourth intermediate value and the target weight is determined as the fifth intermediate value, and the difference between the third intermediate value and the fifth intermediate value is determined as the minimum value of the second interval; The production scenario to which the production data belongs on the day preceding the historical production data is defined as the third production scenario, and the standard deviation of the product output observations in the third production scenario is defined as the third standard deviation. The production scenario to which the production data belongs on the day following the historical production data is defined as the fourth production scenario, and the standard deviation of the product output observations in the fourth production scenario is defined as the fourth standard deviation. The sum of the third and fourth standard deviations is defined as the sixth median. The product of the sixth median and the target weight is defined as the seventh median. The sum of the seventh median and the third median is defined as the maximum value of the second interval.

7. A coking product yield correction system, characterized in that, include: The data acquisition module is used to acquire historical production data of coking products. The historical production data includes production data for multiple consecutive days, and the production data includes: production condition data and product output observations. The scenario segmentation module is used to segment all the production data into scenarios based on the production condition data and preset scenario classification rules, so as to determine the production scenario to which the production data of each day belongs. The prediction model building module is used to traverse the product output observations under each production scenario to build a probability density prediction model. The probability density prediction model corresponds one-to-one with the production scenario and describes the probability distribution of the output of coking products under the corresponding production scenario. The correction module is used to construct an optimization model based on the probability density prediction model, and to use the optimization model to complete the correction of coking product output. The scenario classification rule setting module is used to filter target data corresponding to preset production condition indicators from all production condition data, wherein there is at least one production condition indicator, and all target data corresponding to one production condition indicator constitute a target data group; by dividing all target data groups into intervals, at least one interval segment is obtained for each target data group; an interval segment is selected from each target data group and combined to obtain the production scenario, wherein the production scenario includes a scenario number and an interval segment, and the interval segments in the production scenario correspond one-to-one with the production condition indicators; and the scenario classification rule is determined based on all the obtained production scenarios. The optimization model includes: an objective function, a first constraint, and a second constraint. The objective function is used to obtain the target solution of the decision variables of the probability density prediction model. The decision variables refer to the correction value variables of the output of coking products. The target solution is obtained by determining the value of the corresponding decision variable as the target solution when the product of probabilities of all decision variables is maximized. The first constraint is used to restrict the value of a single decision variable to a first interval, which is determined based on the mean and standard deviation of the product output observations under the current production scenario. The second constraint is used to restrict the sum of the values ​​of all decision variables to a second interval, which is determined based on the sum of the product output observations over multiple consecutive days. The correction module is specifically used to obtain a correction value for the output of coking products over several consecutive days based on the objective function, the first constraint, and the second constraint. The correction value is the corrected output value of the coking products.

8. An electronic device, characterized in that, It includes a processor, a memory, and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement the coking product yield correction method as described in any one of claims 1 to 6.

Images