Method for controlling coal consumption system

Abstract

The present invention provides a method for controlling a coal consumption system, wherein the coal consumption system includes a central control system and multiple monitoring points, wherein the multiple monitoring points are arranged on multiple production line equipment. The method for controlling a coal consumption system includes the control points obtaining monitoring data, the central control system selecting a candidate feature set and establishing a coal consumption model based on the monitoring data, the central control system generating recommended operating values ​​for controllable monitoring points among the multiple monitoring points based on the coal consumption model, and the production line equipment operating according to the recommended operating values ​​to achieve the target operating conditions. The method for controlling a coal consumption system of the present invention can greatly shorten the adjustment interval of production line equipment, improve adjustment efficiency and accuracy, reduce the operating difficulty of operators, and effectively reduce coal consumption.

Description

Technical Field

[0001] The present invention relates to a method for controlling coal consumption, and in particular to a method for controlling a coal consumption system. Background Art

[0002] Cement production consumes coal, but excessive coal consumption not only increases manufacturing costs but also causes environmental pollution. Cement production lines consist of numerous pieces of equipment, each with nearly 10,000 monitoring points. While these monitoring points provide data on kiln conditions, the sheer number of points makes it difficult for operators to comprehensively and quickly assess kiln conditions based on data from all of them. This results in long intervals between equipment adjustments, making adjustments less timely and accurate, and thus failing to effectively reduce coal consumption. Summary of the Invention

[0003] The present invention provides a method for controlling a coal consumption system. The coal consumption system includes a central control system and multiple monitoring points, each of which is located on multiple production line equipment. The method includes obtaining monitoring data from the monitoring points, selecting a candidate feature set and establishing a coal consumption model based on the monitoring data, generating recommended operating values ​​for controllable monitoring points from the multiple monitoring points based on the coal consumption model, and operating the production line equipment based on the recommended operating values ​​to achieve target operating conditions.

[0004] The method for controlling the coal consumption system of the present invention can greatly shorten the adjustment interval of production line equipment, improve the adjustment efficiency and accuracy, reduce the operating difficulty of operators, and effectively reduce coal consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0006] Figure 1 Schematic diagram of a coal consumption system in an embodiment of the present invention.

[0007] Figure 2 for Figure 1 Flowchart of the method of operation of the medium coal consumption system.

[0008] Figure 3 for Figure 2 Flowchart of step S204 of the operation method.

[0009] Figure 4 for Figure 3 Flowchart of modeling step S306.

[0010] Figure 5 for Figure 3 Another flowchart of the modeling step S306. DETAILED DESCRIPTION

[0011] Figure 1 This is a schematic diagram of a coal consumption system 1 in an embodiment of the present invention. The coal consumption system 1 can be used in cement plants to reduce coal consumption. Coal consumption can be expressed as standard coal consumption. The coal consumption system 1 includes monitoring points 101 to 10N and a central control system 12, where N is a positive integer greater than 1. The central control system 12 can run in a computer or a computer network. The monitoring points 101 to 10N can be set on multiple production line equipment in the cement plant to measure the monitoring data of the production line equipment, and coupled to the central control system 12 through a wired or wireless connection. The central control system 12 can be set in the cement plant or a remote computer room, and at least one monitoring point can be set on each production line equipment. The monitoring points 101 to 10N can transmit the monitoring data to the central control system 12, and the central control system 12 can store the monitoring data in a database, generate recommended operating values ​​for the production line equipment based on the monitoring data, and transmit the recommended operating values ​​back to the cement plant so that the production line equipment can operate according to the recommended operating values, achieving the benefits of reducing coal and carbon while maintaining stable operation.

[0012] The various production line equipment in a cement plant may include raw material scales, pulverized coal scales, preheaters, precalciners, rotary kilns, grate coolers, high-temperature fans, cooling fans, kiln head hoods, kiln head exhaust fans, and other production line equipment. The pulverized coal scales feed pulverized coal into the rotary kiln and the precalciner at the kiln tail for combustion. The resulting hot air flow preheats the cement raw material fed from the raw material scales into the preheater. The raw material then enters the precalciner, passes through the rotary kiln, and is cooled by the grate cooler to produce cement clinker. The production of cement clinker can be affected by the temperature and pressure at the preheater outlet, the temperature and pressure at the precalciner inlet, the speed and temperature of the rotary kiln, the air volume and temperature of the high-temperature and cooling fans, and the amount of coal. Monitoring points 101 to 10N may be speedometers, accelerometers, tachometers, air flow meters, current sensors, voltage sensors, power sensors, temperature sensors, pressure sensors, vibration sensors, concentration sensors, image sensors, audio sensors, or other sensors. Monitoring points 101 to 10N can be categorized as controllable monitoring points, uncontrollable monitoring points, environmental monitoring points, quality monitoring points, structured data monitoring points, and unstructured data monitoring points. Controllable monitoring points can monitor controllable variables such as speed, air volume, current, voltage, and power. Uncontrollable monitoring points can monitor uncontrollable variables such as temperature and pressure. Environmental monitoring points can monitor environmental variables such as nitrogen oxide concentration or carbon monoxide concentration. For example, concentration sensors can monitor nitrogen oxide (NOx) concentration and / or carbon monoxide (CO) concentration to ensure that the coal consumption model's recommendations do not exceed environmental quality standards. Quality monitoring points can monitor quality variables such as cement raw material quantity or cement clinker quantity to ensure that the quality of cement produced is acceptable. Structured data monitoring points can monitor variables that can be recorded in a list format, such as speed, air volume, temperature, pressure, concentration, and vibration. Unstructured monitoring points can monitor variables that cannot be recorded in a list format, such as images or sound.

[0013] Monitoring points 101 to 10N can be set at key monitoring locations of production line equipment to monitor variables that affect the production of cement clinker. For example, the inlet of the decomposition furnace is a key monitoring location. The temperature and pressure at the inlet of the decomposition furnace can reflect the state of the cement raw material. Therefore, temperature sensors and pressure sensors can be set at the inlet of the decomposition furnace to sense the temperature and pressure at the inlet of the decomposition furnace. The central control system 12 can generate recommended operating values ​​for the amount of cement raw material and coal consumption based on the temperature and pressure at the inlet of the decomposition furnace. In some embodiments, the central control system 12 can set the recommended operating values ​​for the amount of cement raw material and coal consumption based on empirical values, thereby reducing the coal consumption of the cement plant by 0.28% (approximately 7,000 metric tons). The central control system 12 can transmit the recommended operating values ​​for the amount of cement raw material and coal consumption to the display equipment of the cement plant. The operator can set the raw material scale and coal powder scale according to the recommended operating values ​​displayed on the display screen. In some embodiments, the central control system 12 may also transmit the recommended operating values ​​of cement raw meal quantity and coal consumption to the raw meal scale and coal powder scale respectively, and the raw meal scale and coal powder scale may automatically adjust the cement raw meal quantity and coal consumption according to the recommended operating values ​​respectively.

[0014] In addition, a tachometer and a thermometer can be installed on the rotary kiln, a pressure sensor and a thermometer can be installed on the kiln head hood, a tachometer can be installed on the kiln head exhaust fan, a current sensor and an air volume meter can be installed on the high-temperature fan and the cooling fan, a speed meter can be installed on the grate cooler, and other monitoring points can be installed at other key monitoring locations of the production line equipment. In some embodiments, the central control system 12 can establish a coal consumption model based on the monitoring data measured by the monitoring points 101 to 10N, and generate recommended operating values ​​for the controllable monitoring points among the monitoring points 101 to 10N based on the coal consumption model, thereby further reducing the coal consumption of the cement plant by 1.02% (approximately 24,000 metric tons). For example, the controllable monitoring points can be the tachometer of the rotary kiln and the air volume meter of the cooling fan. The central control system 12 can generate recommended operating values ​​for the speed of the rotary kiln and the air volume of the cooling fan based on the coal consumption model. The central control system 12 can transmit recommended operating values ​​for the rotary kiln speed and cooling fan air volume to a display device at the cement plant. Operators can then set the rotary kiln and cooling fan according to the recommended operating values ​​displayed on the display screen. In some embodiments, the central control system 12 can also transmit the recommended operating values ​​for the rotary kiln speed and cooling fan air volume to the rotary kiln and cooling fan, respectively. The rotary kiln and cooling fan can then automatically adjust the amount of cement raw material and coal consumption, respectively, based on the recommended operating values.

[0015] In some embodiments, the central control system 12 may default to using the coal consumption model to generate recommended operating values. If the recommended operating values ​​exceed upper and lower limits, or if operating the production line equipment based on the recommended operating values ​​would cause a critical operating condition, the central control system 12 may switch to generating recommended operating values ​​based on empirical data or manually determining the recommended operating values. Critical operating conditions may include temperature exceeding upper and lower limits, pressure exceeding upper and lower limits, or rotary kiln speed exceeding upper and lower limits.

[0016] Figure 2 This is a flow chart of an operating method 200 for coal consumption system 1. Operating method 200 includes steps S202 through S208, which are used to establish a coal consumption model and use the model to generate recommended operating values, thereby enabling the cement plant's production line equipment to achieve target operating conditions. Any reasonable technical changes or adjustments to these steps fall within the scope of the present invention. Details of steps S202 through S208 are as follows:

[0017] Step S202: Monitoring points 101 to 10N obtain monitoring data;

[0018] Step S204: The central control system 12 selects a candidate feature set and establishes a coal consumption model based on the monitoring data;

[0019] Step S206: The central control system 12 generates recommended operating values ​​for controllable monitoring points among the monitoring points 101 to 10N according to the coal consumption model;

[0020] Step S208: The production line equipment operates according to the recommended operating values ​​to achieve the target operating conditions.

[0021] In step S202, monitoring points 101 to 10N obtain monitoring data at the same predetermined time interval. The predetermined time interval may be 30 seconds. In step S204, the central control system 12 selects a candidate feature set and establishes a coal consumption model. The candidate feature set may include multiple candidate features, such as cement raw material quantity, cement clinker quantity, coal consumption, temperature of the decomposition furnace inlet, rotation speed of the rotary kiln, and air volume of the cooling fan. The central control system 12 uses the monitoring data corresponding to the candidate features to establish a coal consumption model, and determines whether to update the candidate features in the candidate feature set based on the prediction error generated by the coal consumption model. The detailed steps of step S204 are shown in the attached figure. Figure 3, which will be explained in subsequent paragraphs. In step S206, the central control system 12 generates recommended operating values ​​for controllable monitoring points at fixed times or not at fixed times. The fixed time may be 10 minutes. The central control system 12 may adjust the monitoring data of each controllable monitoring point within its upper and lower limits to provide recommended operating values ​​for each controllable monitoring point to reduce coal consumption. For example, if the current amount of raw cement is 0.5 metric tons and the upper and lower limits of the amount of raw cement are plus or minus 2 metric tons per hour, the central control system 12 may recommend increasing the recommended operating value of the amount of raw cement to 1 metric ton to reduce coal consumption. The upper and lower limits of the monitoring data of each controllable monitoring point may be obtained from empirical values ​​or calculated from a coal consumption model or other models when a stable operating condition is achieved. Other models may be pressure models or temperature models. Stable operating conditions may include maintaining the temperature within the upper and lower temperature limits and maintaining the pressure within the upper and lower pressure limits. For example, if the upper and lower limits of the temperature are 1400°C to 1200°C, the central control system 12 can adjust the amount of cement raw material in the temperature model to determine that the upper and lower limits of the amount of cement raw material must be maintained between plus or minus 2 metric tons per hour in order to maintain the temperature between 1400°C and 1200°C. In some embodiments, if the number of controllable monitoring points is greater than 1, the central control system 12 can set a priority for the recommended operating value of each controllable monitoring point or determine the priority based on the importance of the impact of each controllable monitoring point on coal consumption calculated by the coal consumption model, and produce the recommended operating value of the production line equipment based on the priority. The operator can use the recommended operating value to set the production line equipment according to the priority, or the production line equipment can operate using the recommended operating value according to the priority. In step S208, the target operating condition can be stable operating conditions, reduced coal consumption, increased production, or any two thereof.

[0022] Figure 3 for Figure 2 A flowchart of step S204 of the operation method is provided. Step S204 includes steps S302 to S312. Steps S302 to S306 are used to establish a coal consumption model, and steps S308 to S312 are used to select a candidate feature set. Any reasonable technical changes or step adjustments fall within the scope of the present invention. The details of steps S302 to S312 are as follows:

[0023] Step S302: The central control system 12 receives a plurality of candidate features as a candidate feature set;

[0024] Step S304: The central control system 12 updates the multiple modeling data sets using the monitoring data according to the candidate feature sets;

[0025] Step S306: The central control system 12 establishes a coal consumption model based on the multiple sets of modeling data sets;

[0026] Step S308: The central control system 12 inputs the updated data in each set of modeling data sets into the coal consumption model to generate a prediction error;

[0027] Step S310: The central control system 12 determines whether the prediction error is greater than the error threshold; if so, proceed to step S312; if not, proceed to step S206;

[0028] Step S312: The central control system 12 updates the candidate feature set.

[0029] In step S302, each candidate feature corresponds to one of monitoring points 101 to 10N. For example, the amount of cement raw meal can correspond to a raw meal scale; the amount of cement clinker can correspond to a clinker scale; the coal consumption can correspond to a pulverized coal scale; the temperature at the precalciner inlet can correspond to a precalciner inlet thermometer; the rotary kiln speed can correspond to a rotary kiln speed meter; and the cooling fan air volume can correspond to a cooling fan air volume meter. In step S304, each modeling dataset corresponds to one of the candidate features, and the central control system 12 can sequentially update the monitoring data of monitoring points 101 to 10N into the corresponding modeling dataset. For example, the central control system 12 can store all cement raw meal amounts in the first modeling dataset, all cement clinker amounts in the second modeling dataset, all precalciner inlet temperature values ​​in the third modeling dataset, all rotary kiln speed values ​​in the fourth modeling dataset, and all cooling fan air volume values ​​in the fifth modeling dataset; and all coal consumption values ​​in the target coal consumption dataset. In step S306, the central control system 12 trains a coal consumption model based on the multiple sets of modeling data sets and the multiple corresponding coal consumption values. For example, the central control system 12 may randomly extract the cement raw material amount, cement clinker amount, decomposition furnace inlet temperature, rotary kiln speed, and cooling fan air volume corresponding to the same time point from the first to fifth sets of modeling data sets and input them into the coal consumption model to calculate the predicted coal consumption value, and adjust the multiple weights in the coal consumption model so that the predicted coal consumption value approaches the coal consumption value corresponding to the same time point, thereby establishing the coal consumption model. After establishing the coal consumption model, the central control system 12 may adjust the priority of the recommended operating values ​​for the corresponding controllable monitoring points based on the importance of the candidate features. The importance of the candidate features is related to the weight of the candidate features calculated in the coal consumption model or set based on professional knowledge. The greater the weight of the candidate feature, the higher the priority of the candidate feature. The smaller the weight of the candidate feature, the lower the priority of the candidate feature. The algorithm for establishing the coal consumption model may include a rule-based algorithm based on domain knowledge, a linear / nonlinear algorithm, an ensemble algorithm, a bagging algorithm, a boosting algorithm, an adaptive learning algorithm, other machine learning algorithms, or a combination of the above algorithms. Figure 4 The hierarchical modeling implementation shown, or by Figure 5The restricted modeling implementation shown is explained in the subsequent paragraphs.

[0030] In step S308, the central control system 12 inputs the updated data from each modeling dataset into the coal consumption model to generate a predicted coal consumption value, and calculates the difference between the predicted coal consumption value and the updated coal consumption value as the prediction error. A smaller prediction error indicates that the coal consumption model is accurate, and the predicted coal consumption value predicted by the coal consumption model is close to the actual updated coal consumption value. A larger prediction error indicates that the coal consumption model is inaccurate, and the predicted coal consumption value predicted by the coal consumption model deviates from the actual updated coal consumption value. In step S310, if the prediction error is greater than the error threshold, the candidate feature set needs to be updated (step S312) to retrain the coal consumption model; if the prediction error is not greater than the error threshold, the candidate feature set does not need to be updated, and the trained coal consumption model can be used in subsequent procedures. In some embodiments, the central control system 12 may remove the candidate feature from the candidate feature set to update the candidate feature set. In other embodiments, the central control system 12 may update another candidate feature to the candidate feature set, where the other candidate feature corresponds to another monitoring point. For example, the other candidate feature may be the pressure of the kiln head hood, and the other monitoring point may be the pressure gauge of the kiln head hood. The other monitoring point can be a previously installed but unused monitoring point on the production line equipment, or a newly installed monitoring point on the production line equipment. For example, a vibration sensor can be added to the rotary kiln to increase the monitoring point.

[0031] Figure 4 This is a flow chart of modeling step S306 in Appendix 3. Modeling step S306 includes steps S402 and S404, which are used to perform hierarchical modeling to generate a coal consumption model. Any reasonable technical changes or step adjustments fall within the scope of the present invention. The details of steps S402 and S404 are as follows:

[0032] Step S402: training a first model based on a corresponding modeling data set of corresponding candidate features of the uncontrollable monitoring point to reduce the residual between the target coal consumption value and the predicted coal consumption value;

[0033] Step S404: training a second model based on the corresponding modeling data set of the corresponding candidate features of the controllable monitoring points to reduce the difference between the residual and the predicted residual value.

[0034] The coal consumption model generated according to the hierarchical modeling includes a first model and a second model, the first model only includes the corresponding candidate features of the uncontrollable monitoring points, and the second model only includes the corresponding candidate features of the controllable monitoring points. In step S402, since the first model only includes the corresponding candidate features of the uncontrollable monitoring points, the residual generated by the trained first model has removed the influence of the uncontrollable monitoring points. In step S404, since the second model only includes the corresponding candidate features of the controllable monitoring points, the trained second model can be used to predict the effect of changes in the controllable monitoring points. In some embodiments, the central control system 12 can adjust the monitoring data of each controllable monitoring point within its upper and lower limits when a stable operating condition is achieved to provide a recommended operating value for each controllable monitoring point, thereby reducing the predicted coal consumption value.

[0035] Figure 5 for Figure 3 Another flow chart of modeling step S306 includes steps S502 and S504 for performing constrained modeling to generate a coal consumption model. Any reasonable technical changes or step adjustments fall within the scope of the present invention. The details of steps S502 and S504 are as follows:

[0036] Step S502: training a coal consumption model based on the multiple sets of corresponding modeling data sets corresponding to the candidate features of the monitoring points 101 to 10N;

[0037] Step S504: Calculate the upper limit and lower limit of the monitoring data of the monitoring point 101n among the monitoring points 101 to 10N using the coal consumption model according to the target output.

[0038] In step S504, the target output is the predetermined target output of cement clinker. The central control system 12 can fix the data of the candidate features of cement clinker in the trained coal consumption model to the predetermined target output and adjust the monitoring data of each controllable monitoring point to obtain the upper and lower limits of the monitoring data of each controllable monitoring point when a stable operating condition is achieved.

[0039] Figures 1 to 5 The embodiment selects a candidate feature set and establishes a coal consumption model based on the monitoring data of the production line equipment, and uses the coal consumption model to provide recommended operating values ​​for the production line equipment to reduce coal consumption, reduce the operating costs of the cement plant, and reduce carbon emissions. The production line equipment can be quickly and correctly controlled to produce cement clinker without manual judgment of the operating values ​​of the production line equipment.

[0040] The above description is only a preferred embodiment of the present invention. All equivalent changes and modifications made according to the claims of the present invention should fall within the scope of the present invention.

Claims

1. A method for controlling a coal consumption system, characterized in that: The coal consumption system includes a central control system and multiple monitoring points, wherein the monitoring points are set on multiple production line equipment. The method includes: The monitoring point obtains monitoring data; The central control system receives a plurality of candidate features as a candidate feature set, each candidate feature corresponding to one of the plurality of monitoring points; The central control system updates multiple sets of modeling data sets corresponding to the multiple candidate features using the monitoring data according to the candidate feature set, each set of modeling data sets corresponding to one of the multiple candidate features; The central control system establishes a coal consumption model based on the multiple sets of modeling data sets; the central control system inputs updated data in each set of modeling data sets into the coal consumption model to generate a prediction error; and The central control system updates the candidate feature set according to the prediction error to retrain the coal consumption model; The central control system generates recommended operating values ​​for controllable monitoring points among the monitoring points according to the established coal consumption model or the retrained coal consumption model; and The production line equipment operates according to the recommended operating values ​​to achieve a target operating condition.

2. The method according to claim 1, wherein The coal consumption system further includes another monitoring point, which is arranged on any production line equipment among the plurality of production line equipment; and The central control system updates the candidate feature set according to the prediction error, including: If the prediction error is greater than an error threshold, the central control system updates another candidate feature to the candidate feature set, where the another candidate feature corresponds to another monitoring point.

3. The method according to claim 1, wherein The central control system updates the candidate feature set according to the prediction error, including: If the prediction error is greater than an error threshold, the central control system removes the corresponding candidate feature from the candidate feature set.

4. The method according to claim 1, wherein The plurality of monitoring points include a plurality of controllable monitoring points and a plurality of uncontrollable monitoring points; The coal consumption model includes a first model and a second model; and The central control system establishes the coal consumption model according to the multiple sets of modeling data sets, including: The central control system trains the first model based on multiple sets of corresponding modeling data sets of multiple corresponding candidate features of the uncontrollable monitoring points to reduce the residual between a target coal consumption value and a predicted coal consumption value; and The central control system trains the second model based on multiple sets of corresponding modeling data sets of multiple corresponding candidate features of the controllable monitoring points to reduce the difference between the residual and a predicted residual value.

5. The method according to claim 1, wherein The central control system establishes or retrains the coal consumption model based on the multiple sets of modeling data sets, including: The central control system trains the coal consumption model based on multiple sets of corresponding modeling data sets of multiple corresponding candidate features of the monitoring points; and The central control system calculates the upper limit and the lower limit of monitoring data of one of the multiple monitoring points using the coal consumption model according to a target output.

6. The method according to claim 5, wherein The central control system also generates a recommended operating value for one of the monitoring points based on the upper limit and the lower limit.

7. The method according to claim 1, wherein A candidate feature in the candidate feature set corresponds to a controllable monitoring point; The method further includes adjusting, by the central control system, the priority of the suggested operation value of the controllable monitoring point according to the importance of the candidate feature; and The central control system generating the recommended operating value for the controllable monitoring point among the multiple monitoring points according to the coal consumption model includes: the central control system generating the recommended operating value according to the priority.

8. The method according to claim 7, wherein The importance is related to the weight of the candidate feature in the coal consumption model.

Images