Intelligent Scheduling Method, Storage Medium and Terminal of Air Compressor Unit Based on Flow Prediction

Abstract

The present invention discloses an intelligent scheduling method, a storage medium and a terminal for an air compressor unit based on flow prediction, belonging to the field of industrial control. The method includes: determining whether the control errors of the flow control method and the pressure control method are within the first threshold range. If so, the flow control method is used to perform scheduling control on the air compressor unit; if not, the pressure control method is used to perform scheduling control on the air compressor unit. The flow control method includes: predicting the flow rate of the air compressor to obtain the predicted flow rate in the future time period; determining the load rates of the industrial frequency machine combination and the variable frequency machine; adjusting the load rate of the variable frequency machine according to the predicted flow rate, so as to achieve the scheduling control of the air compressor unit. According to the present invention, the flow control method adjusts the load rate of the variable frequency machine according to the predicted flow rate, can predict the flow change trend in the next time period in advance, enables the variable frequency machine to work in the high energy efficiency range as much as possible, and reduces the start / stop times and loading / unloading times of the industrial frequency machine, so as to achieve the purpose of energy saving.

Description

Technical Field

[0001] The present invention relates to the field of industrial control, and in particular to an intelligent scheduling method, a storage medium, and a terminal for an air compressor unit based on flow prediction. Background Art

[0002] The world climate is facing increasingly serious problems. Global catastrophic climate changes occur frequently, which have seriously endangered human survival and health. To prevent the further deterioration of the global climate, various countries have taken different measures to save energy, reduce emissions, and reduce carbon emissions. Mine air compressors are important power systems in coal mine production, which consume a large amount of electrical energy. Using various measures to reduce energy consumption is of great value to coal mine production and low-carbon economy.

[0003] Currently, the control strategy of air compressors usually adopts a pressure-based control method. This is mainly because the pressure control cost is low and no additional sensors need to be installed. Secondly, air-powered equipment often has certain pressure requirements, and using pressure control has relatively intuitive system reliability. However, in order for air-powered equipment to operate normally, this control method often needs to compress compressed air to a pressure level 5% - 20% higher than the normal operation of the equipment, which will waste a lot of compressed gas and electrical energy. At the same time, when a system failure or pipeline leakage occurs, pressure control cannot remind users through pressure detection. In the field of compressed air, the compressed gas pressure is one of the important parameters of the control system. Although using pressure control can meet the requirements, it will waste a lot of compressed air and electrical energy.

[0004] In the air compressor air system, the essence of the air compressor control system is a process of dynamic balance of compressed air. Use as much gas as needed and supplement as much gas as used to achieve a dynamic balance state between compressed air systems and ensure the normal operation of air-powered equipment. However, this control method requires monitoring the usage of flow rate. Generally, sensors are used to measure the flow rate in real time, and then the flow rate information is fed back to the control system. However, this control method has a certain lag and cannot provide accurate data reference for subsequent air compressor scheduling control, reducing the reliability of air compressor scheduling. Summary of the Invention

[0005] The purpose of the present invention is to overcome the problems of the prior art, and provide an intelligent scheduling method, a storage medium, and a terminal for an air compressor unit based on flow prediction.

[0006] The purpose of the present invention is achieved by the following technical solutions: An intelligent scheduling method for an air compressor unit based on flow prediction, the method includes the following steps:

[0007] Judge whether the control errors of the flow control method and the pressure control method are within the first threshold range. If so, use the flow control method to perform scheduling control on the air compressor unit; if not, use the pressure control method to perform scheduling control on the air compressor unit;

[0008] The flow control method includes:

[0009] Predict the flow rate of the air compressor to obtain the predicted flow rate for the future time period;

[0010] Determine the load rates of the industrial frequency machine combination and the variable frequency machine;

[0011] Adjust the load rate of the variable frequency machine according to the predicted flow rate, so as to realize the scheduling control of the air compressor unit.

[0012] In one example, after determining the operating state of the industrial frequency machine combination and the load rate of the variable frequency machine, it further includes:

[0013] Introduce the current pressure value and determine the operating state of the industrial frequency machine combination and the load rate of the variable frequency machine again.

[0014] In one example, the pressure control method includes:

[0015] Obtain the pressure value in the pipeline and judge whether the pressure value exceeds the second threshold;

[0016] If it exceeds the lower limit of the second threshold, judge whether the variable frequency machine is fully loaded. If so, start at least one industrial frequency machine according to the full load duration of the variable frequency machine;

[0017] If it exceeds the upper limit of the second threshold, judge whether the variable frequency machine is unloaded. If so, shut down at least one industrial frequency machine according to the unload duration of the variable frequency machine.

[0018] In one example, when adjusting the load rate of the variable frequency machine according to the predicted flow rate, the load rate of the variable frequency machine is made to be between 60% and 80%.

[0019] In one example, the specific prediction of the flow rate of the air compressor includes:

[0020] Divide the original air compressor flow rate data sequence into a horizontal sequence and a vertical sequence: the original air compressor flow rate data sequence is the horizontal sequence, extract the data in the original air compressor flow rate data sequence at intervals to obtain the horizontal sequence, and continue to extract the data in the horizontal sequence and / or the original air compressor flow rate data sequence at intervals to obtain more horizontal sequences; the relationship between the horizontal sequences is the vertical sequence;

[0021] Input the horizontal sequence into the first neural network model for flow rate prediction to obtain the first prediction result; input the vertical sequence into the second neural network model for flow rate prediction to obtain the second prediction result;

[0022] Perform mean processing or weighted processing on the first prediction result and the second prediction result to obtain the predicted flow rate of the air compressor in the future time period.

[0023] In one example, when inputting the horizontal sequence into the ARIMA model for flow rate prediction, it further includes:

[0024] Conduct a stationarity test on the horizontal sequence.

[0025] In one example, before dividing the original air compressor flow rate data into horizontal and vertical sequences, it further includes:

[0026] Conduct data compensation processing on the original air compressor flow rate data sequence: when there are short-term data missing in the original air compressor flow rate data sequence, compensate according to the mean value of the adjacent data of the missing data;

[0027] When there are long-term data missing in the original air compressor flow rate data sequence, compensate according to the historical data similarity.

[0028] In one example, before inputting the horizontal sequence or the vertical sequence into the neural network model, it further includes:

[0029] Conduct Kalman filtering processing on the horizontal sequence and / or the vertical sequence.

[0030] It should be further noted that the technical features corresponding to the above examples can be combined or replaced with each other to form a new technical solution.

[0031] The present invention further includes a storage medium, on which computer instructions are stored, and when the computer instructions run, they execute the steps of the intelligent scheduling method for the air compressor unit based on flow rate prediction formed by any one of the above examples or a combination of multiple examples.

[0032] The present invention further includes a terminal, including a memory and a processor, where computer instructions that can run on the processor are stored on the memory, and when the processor runs the computer instructions, it executes the steps of the intelligent scheduling method for the air compressor unit based on flow rate prediction formed by any one of the above examples or a combination of multiple examples.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] 1. In one example, the flow control method of the present invention adjusts the load rate of the variable-frequency machine according to the predicted flow rate, eliminates the hysteresis effect caused by the existing flow detection, can predict the flow rate change trend in the next time period in advance, replenishes compressed air into the gas storage tank in advance, ensures the supply pressure, reduces the specific power of the air compressor unit at the same time, enables the variable-frequency machine to work in the high-efficiency range as much as possible, reduces the start / stop and loading / unloading times of the power-frequency machine, reduces the energy consumption of the air compressor, and achieves the purpose of energy saving; further, the flow control method and the pressure control method are run simultaneously, and the current control strategy is determined according to the error between the two, ensuring the reliability of the flow control method when it is selected for control, and since the flow control method eliminates the hysteresis effect, it is beneficial to make accurate scheduling decisions in advance, so as to realize intelligent scheduling.

[0035] 2. In one example, after the longitudinal sequence is extracted at intervals through the transverse sequence, it is a sequence with a longer time span, and this group of sequences contains more periodic regular data. Then, the air compressor flow rate predicted by using the longitudinal sequence (the relationship between the transverse sequences) has the characteristics of periodic regularity, so as to improve the accuracy of the final flow rate prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The following further describes in detail the specific embodiments of the present invention with reference to the accompanying drawings. The accompanying drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The same reference numerals are used to represent the same or similar parts in these drawings. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application.

[0037] Figure 1 It is a flowchart of the intelligent scheduling method for the air compressor unit in one example of the present invention;

[0038] Figure 2 It is a performance curve diagram of the specific power of the variable-frequency machine;

[0039] Figure 3 It is a flowchart of the flow control method in one example of the present invention;

[0040] Figure 4 It is a flowchart of the intelligent scheduling method for the air compressor unit in another example of the present invention;

[0041] Figure 5 It is a flowchart of the combined scheduling method for the air compressor unit in one example of the present invention;

[0042] Figure 6 It is a flowchart of the pressure control method in one example of the present invention;

[0043] Figure 7 It is a schematic diagram of the transverse sequence and the longitudinal sequence in one example of the present invention;

[0044] Figure 8 This is the prediction result graph of the ARIMA model in an example of the present invention;

[0045] Figure 9 This is the flowchart of the long-term data missing compensation method in an example of the present invention;

[0046] Figure 10 This is the schematic diagram of the partial flow compensation result of the air compressor unit at a certain coal mine site in an example of the present invention;

[0047] Figure 11 This is the schematic diagram of the result after performing Kalman filtering on some horizontal sequences and vertical sequences in an example of the present invention;

[0048] Figure 12 This is the flowchart of the intelligent scheduling method for the air compressor unit in a preferred example of the present invention. Specific embodiments

[0049] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0050] In the description of the present invention, it should be noted that the directions or positional relationships indicated by terms such as "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be construed as a limitation of the present invention. In addition, the use of ordinal numbers (for example, "first and second", "first to fourth", etc.) is to distinguish objects and is not limited to this order, and cannot be construed as indicating or implying relative importance.

[0051] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "installation", "connection", and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0052] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0053] In one example, asFigure 1 As shown in the figure, an intelligent scheduling method for air compressor units based on flow prediction specifically includes the following steps:

[0054] S1’: Determine whether the control errors of the flow control method and the pressure control method are within the first threshold range; if so, proceed to step S2’; if not, proceed to step S3’; where the first threshold range is obtained from historical experience or experimental results.

[0055] S2’: Use the flow control method to perform scheduling control on the air compressor unit;

[0056] S3’: Use the pressure control method to perform scheduling control on the air compressor unit.

[0057] In this example, according to the results of the intelligent scheduling algorithm (flow control method or pressure control method), the status of each air compressor is sent to the corresponding air compressor. At the same time, the control center further determines whether the predicted air compressor combination (the combination of the working variable-frequency machine and the industrial-frequency machine) meets the requirements; otherwise, the previous combination is used to ensure production operation. Among them, the air compressor includes a variable-frequency (air compressor) machine and an industrial-frequency (air compressor) machine, and the air compressor unit is the combination of the variable-frequency machine and the industrial-frequency machine. The intelligent scheduling control of the air compressor unit includes the specific power control of the variable-frequency machine and the working state control of the industrial-frequency machine.

[0058] Specifically, the specific power of the variable-frequency air compressor is not constant but a curve. In order to reduce energy consumption, the variable-frequency air compressor should work in the region with a relatively high specific power as much as possible. The specific power calculation formula is as follows:

[0059]

[0060] In the formula, P B represents the specific power, P Z represents the air compressor power, and Q represents the flow rate. In the actual control process, after the variable-frequency machine starts, it should work in the region with a relatively low specific power as much as possible. Preferably, the load rate of the variable-frequency machine is in the range of 60% - 80%, so that the air compressor can discharge more compressed air under the same energy consumption. Among them, the specific power performance curve of the variable-frequency machine is as Figure 2 shown.

[0061] For the industrial-frequency machine, since its specific power is fixed, more electric energy will be wasted during startup and shutdown, loading and unloading. Therefore, by controlling and reducing the frequent startup and shutdown times, loading and unloading times of the industrial-frequency machine, the purpose of energy saving for the industrial-frequency machine is achieved.

[0062] Furthermore, as Figure 3 shown, the flow control method includes the following steps:

[0063] S21’: Predict the air compressor flow rate to obtain the predicted flow rate for the future time period. As an option, the air compressor flow rate can be predicted through a neural network model, including but not limited to ARIMA model, LSTM model, GRU model, BP neural network, etc.

[0064] S22’: Determine the load rates of the industrial frequency machine combination and the variable frequency machine. Among them, the industrial frequency machine combination is the set of all industrial frequency machines that need to work currently. The industrial frequency machine scheduling control is achieved by controlling the start, stop, loading, and unloading states of each industrial frequency machine. The information of the industrial frequency machine combination and the load rate information of the variable frequency machine can be fed back through on-line detection equipment.

[0065] S23’: Adjust the load rate of the variable frequency machine according to the predicted flow rate to make the load rate of the variable frequency machine be between 60% and 80%, so as to achieve the scheduling control of the air compressor unit.

[0066] Further, in this example, the pressure control method is an existing control algorithm, that is, the compressed air is compressed to a pressure level that is 5% - 20% higher than the normal operation of the variable frequency machine and the industrial frequency machine.

[0067] In this example, both the flow control method and the pressure control method are adopted, and the current control strategy is determined according to the error between the two. Among them, the pressure control method has been widely used and has high reliability. Then, when the error between the operation of the flow control method and the pressure control method is within the first error range, the reliability of the flow control method can also be guaranteed. Moreover, since the flow control method eliminates the hysteresis effect, it is conducive to making accurate scheduling decisions in advance, so as to achieve intelligent scheduling. Further, in this example, the flow control method adjusts the load rate of the variable frequency machine according to the predicted flow rate, eliminates the hysteresis effect caused by the existing flow detection, can predict the flow change trend in the next time period in advance, replenish compressed air into the gas storage tank in advance, ensure the supply air pressure, reduce the specific power of the air compressor unit at the same time, make the variable frequency machine work in the high energy efficiency range as much as possible, and reduce the loading / unloading and start / stop times of the industrial frequency machine, so as to reduce the energy consumption of the air compressor and achieve the purpose of energy saving.

[0068] In one example, after determining the working state of the industrial frequency machine combination and the load rate of the variable frequency machine, it further includes:

[0069] Introduce the current pressure value and determine the working state of the industrial frequency machine combination and the load rate of the variable frequency machine again. At this time, the intelligent scheduling method of the air compressor unit (the set of all variable frequency machines that need to work currently) is as Figure 4 shown, and the combined scheduling method of the air compressor unit is as Figure 5 shown. This example introduces real-time pressure value for compensation to ensure that while accurately controlling the flow rate, the air compressor system is at the lowest pressure, reducing energy consumption waste, and effectively ensuring that pneumatic equipment can be used normally.

[0070] In one example, as Figure 6As shown, the pressure control method includes:

[0071] Obtain the main pipe pressure value and determine whether the pressure value exceeds the second threshold (set value);

[0072] If it exceeds the lower limit of the second threshold, determine whether the frequency converter is fully loaded. If so, start at least one industrial frequency machine according to the full-load duration of the frequency converter; Preferably, when the load of the frequency converter is greater than 95%, it is determined that it is fully loaded; Preferably, when the full-load duration of the frequency converter is greater than 10 s, start the first industrial frequency machine; If the full-load duration of the frequency converter is still greater than 10 s, continue to start other industrial frequency machines.

[0073] If it exceeds the upper limit of the second threshold, determine whether the frequency converter is unloaded. If so, turn off at least one industrial frequency machine according to the unload duration of the frequency converter. Preferably, when the load of the frequency converter is less than 35%, it is determined that it is unloaded; Preferably, when the unload duration of the frequency converter is greater than 10 s, turn off the first industrial frequency machine; If the unload duration of the frequency converter is still greater than 10 s, continue to turn off other industrial frequency machines. In practical applications, compressed air pneumatic equipment has certain pressure requirements for compressed air. When the pressure is less than this value, the equipment cannot be used normally. In order to ensure that the pneumatic equipment can be used normally, it is necessary to control the minimum pressure value of the air storage tank, and the pressure control method can meet the minimum pressure requirement and has high stability. Therefore, the pressure control method is also often used for air compressor control.

[0074] In an example, the specific prediction of the air compressor flow rate includes:

[0075] S1”: Divide the original air compressor flow data sequence into a horizontal sequence and a vertical sequence: The original air compressor flow data sequence is the horizontal sequence. Intervals are used to extract data from the original air compressor flow data sequence to obtain the horizontal sequence, and continue to extract data from the horizontal sequence and / or the original air compressor flow data sequence at intervals to obtain more horizontal sequences. Preferably, continue to extract data from the newly generated horizontal sequences in turn at intervals to obtain more horizontal sequences; The relationship between the horizontal sequences is the vertical sequence;

[0076] S2”: Input the horizontal sequence into the first neural network model for flow rate prediction to obtain the first prediction result; Input the vertical sequence into the second neural network model for flow rate prediction to obtain the second prediction result;

[0077] S3”: Perform mean processing or weight processing on the first prediction result and the second prediction result to obtain the predicted flow rate of the air compressor in the future time period.

[0078] Specifically, during the process of dividing the horizontal sequence in step S1”, each interval is preferably a different interval, and of course, it can also be an equal interval.

[0079] As an option, such as Figure 7As shown, the sequence division method in step S1” can be as follows: The original data is a set of horizontal sequences, denoted as group ①. Every 6 data are extracted once to obtain another set of horizontal sequences, denoted as group ②. Every 4 data in group ② are extracted once to obtain horizontal sequence group ③, and so on, until 5 groups of horizontal sequences are finally obtained; the relationship between the horizontal sequences is the vertical sequence.

[0080] As an option, the sequence division method in step S1” can also be as follows: The original air compressor flow data is a set of horizontal sequences, denoted as group ①; then every 6 data are extracted once and the mean value is calculated to form another set of horizontal sequences, denoted as group ②; every 4 data in group ② are extracted once and the mean value is calculated to obtain horizontal sequence group ③; every 5 data in group ③ are extracted once and the mean value is calculated to obtain horizontal sequence ④; every 4 data in group ④ are extracted once and the mean value is calculated to obtain horizontal sequence ⑤; every 3 data in group ⑤ are extracted once and the mean value is calculated to obtain horizontal sequence ⑥; the relationship between the horizontal sequences is the vertical sequence.

[0081] Specifically, the first neural network model and the second neural network model need to be models with time series prediction. By analyzing the law of change of a certain variable over time, they predict the future change trend. Time series prediction is often used to predict time series data, such as sales, stock prices, temperature, flow rate, energy consumption, etc. In the field of time series prediction, a large number of time series prediction models have been proposed, such as the ARIMA model, BP neural network, LSTM neural network, etc. The first neural network model is preferably the ARIMA model; the second neural network model is preferably the BP neural network. In order to obtain more accurate and reliable air compressor flow prediction results, the present invention uses a bidirectional combined hybrid prediction model to predict the air compressor flow. Specifically, the horizontal sequence data is substituted into the ARIMA model for prediction to obtain the first prediction result (horizontal prediction result) of each group of horizontal sequences; for vertical sequence prediction, different vertical sequence data need to be brought into the trained BP neural network for prediction; the horizontal prediction result and the vertical prediction result (the second prediction result) both predict the result at the same time. Among them, the horizontal sequence contains the change trend of the sequence. In the ARIMA model, only a small amount of data is required for horizontal prediction. Using a large amount of original horizontal sequence data not only consumes a lot of resources but also cannot improve the prediction accuracy. Therefore, it is difficult to effectively utilize the periodic law of data by using only the original horizontal sequence. After obtaining multiple groups of horizontal sequences by interval extraction, their time span is longer and they also contain more periodic law data. The vertical sequence is the relationship between the horizontal sequences. Since the latter group of horizontal sequences is generated from the previous group of horizontal sequences according to the interval, there is a corresponding relationship between the two sequences. In the present invention, this relationship is regarded as the vertical sequence relationship, and this relationship can be given by the following formula;

[0082] x j = x i p ij

[0083] In the formula, x j represents the data of the previous horizontal sequence, and x i is the data of the subsequent horizontal sequence, and p ij is the relationship between x i and x j This parameter is obtained through training by the BP neural network, and the vertical sequence is the sequence composed of data such as x i and x j Therefore, the data at the same time node can be predicted through the horizontal sequence or the vertical sequence. The horizontal prediction result and the vertical prediction result are averaged to obtain a two-way combined prediction result (the predicted flow rate of the air compressor in the future time period) that combines the advantages of the horizontal prediction result and the vertical prediction result, thereby further improving the prediction accuracy.

[0084] For vertical prediction, the BP neural network is used. Different BP neural networks with different sizes are used between different vertical sequences. After being trained, the BP neural network is used to predict the vertical relationship in the horizontal sequence to ensure that the prediction method can capture the periodic law in the time series.

[0085] Considering that the air compressor flow time series has strong regularity and a large amount of data, in this example, a hybrid prediction method combining horizontal and vertical is used to predict the air compressor flow. This method can also maximize the mining of the law and non-linear relationship of the air compressor flow when dealing with massive time series data, thereby improving the prediction accuracy.

[0086] In one example, when inputting the horizontal sequence into the ARIMA model for flow prediction, it also includes:

[0087] Performing a stationarity test on the horizontal sequence. Specifically, performing stationarity tests and Akaike information criterion tests on multiple groups of horizontal sequence data respectively, selecting appropriate ARIMA model parameters, and substituting the data into different ARIMA models for prediction to obtain the prediction results of each group of horizontal sequences; for vertical sequence prediction, different vertical sequence data need to be substituted into the trained BP neural network for prediction.

[0088] Furthermore, the stationarity test specifically includes:

[0089] In the ARIMA(p, d, q) model, three parameters need to be determined. The parameter d is used to perform differential calculation on the original data. If the time series is stationary, no differencing process is required and its value is 0. When using the ARIMA model for prediction, there are strict requirements for the input data. Therefore, a stationarity test is performed on the input data. The stationarity test method adopted in this method is the Augmented Dickey–Fuller (ADF) test, which is used to determine whether the filtered air compressor flow time series is stationary. The principle of ADF is as follows: Assume that the air compressor flow time series is non-stationary. If the obtained significance test statistic is less than three confidence levels (10%, 5%, and 1%), then there is a corresponding (90%, 95%, and 99%) possibility that the air compressor flow time series is stationary.

[0090] Table 1 ADF test results of the input sequence

[0091]

[0092] In the ADF test results of this example, the significance test value is significantly less than the 1% confidence level. Therefore, there is a 99% probability that the sequence is stationary.

[0093] To select the best parameters in the ARIMA model, this method uses R-squared, S.E. of regression, and the Akaike information criterion (AIC) to evaluate the ARIMA model. The evaluation results are shown in Table 2 as follows:

[0094] Table 2 ARIMA model evaluation results

[0095]

[0096]

[0097] The smaller the function value of AIC, the better the model performance. Since this method requires a large number of attempts in the actual application process, only partial model evaluation results are shown in Table 2. According to the output results in Table 2 to evaluate the ARIMA model, the larger the R-squared (coefficient of determination), the smaller the S.E. Regression (standard error of regression), and the smaller the AIC (Akaike information criterion), the better the model performance. Therefore, in the specific implementation of this method, we finally select ARIMA(1, 0, 2), substitute the data for prediction, and its prediction results are as Figure 8 shown, which can accurately reflect the trend of the flow data.

[0098] In an example, before dividing the original air compressor flow data into horizontal and vertical sequences, it also includes:

[0099] Data compensation processing is performed on the original air compressor flow data sequence. The historical air compressor flow data is sorted out, and the missing data types are identified. Different compensation methods are used for different missing data; the missing data includes short-term missing and long-term missing. Short-term missing is often composed of a few scattered missing data, and long-term missing is usually composed of a section of missing data. There are many reasons for data missing in the air compressor system data acquisition, such as industrial Internet of Things network failures, sensor failures, sensor signal transmission errors, etc. Through statistical analysis, we found that most data missing is caused by equipment instability or signal transmission errors; missing data will reduce the effectiveness of the data, and further affect the results of mathematical statistics and time series prediction. Therefore, it is necessary to compensate for the missing data.

[0100] Specifically, when short-term data missing occurs in the original air compressor flow data sequence, it is compensated according to the mean value of the adjacent data of the missing data, that is, the mean value obtained from the adjacent data is used as the missing data. The specific method is as follows:

[0101]

[0102] Among them, x i represents the data to be compensated, and i represents the position of the data to be compensated.

[0103] When long-term data missing occurs in the original air compressor flow data sequence, it is compensated according to the historical data similarity. As Figure 9 shown, specifically, the database method is used for compensation, that is, similar fluctuation situations are searched for at other times to establish a database. During the compensation process, the database is called and supplemented with a certain random value to avoid repetition. After compensating for the missing data according to the above, some flow compensation results of the air compressor unit at a coal mine site are as Figure 10 shown.

[0104] In an example, before inputting the horizontal sequence or the vertical sequence into the neural network model, it further includes:

[0105] Performing Kalman filtering processing on the horizontal sequence and / or the vertical sequence. Preferably, Kalman filtering processing is performed on both the horizontal sequence and the vertical sequence. Specifically, the data processing formula of Kalman filtering is:

[0106] P k|k-1 = P k-1|k-1 + Q k

[0107]

[0108] x k|k = x k|k-1 + K k (R k - xk|k-1 )

[0109] P k|k =(1 - K k )P k|k-1

[0110] where P k|k-1 is the system covariance matrix at time k; P k-1|k-1 is the system covariance matrix at time k - 1; Q k is the covariance of the system process noise; K k is the Kalman gain, which is an intermediate result of filtering; Z k is the covariance matrix of the object measurement noise; x k|k is the optimal estimated value of the state variable at time k; R k is the measured value of the object; P k|k is the covariance matrix at time k. To obtain stable input data, the Kalman filter is used to process the air compressor flow rate. Among them, the Kalman filter parameters are set as follows: Q k = 0.001, Z k = 0.3, and other default parameters are set to 0. The results of the Kalman filter for some horizontal sequences and vertical sequences are as Figure 11 shown. It can be seen that the processed data is smoother after filtering.

[0111] Now, the above examples are combined to obtain the preferred example of the present invention, as Figure 12 shown. At this time, the scheduling method includes the following steps:

[0112] S1: Data preprocessing: Perform data compensation processing on the original air compressor flow rate data sequence;

[0113] S2: Divide the original air compressor flow rate data sequence into horizontal sequences and vertical sequences;

[0114] S3: Perform Kalman filter processing on the horizontal sequences and vertical sequences;

[0115] S4: Input the horizontal sequence into the ARIMA model for flow rate prediction to obtain the first prediction result; input the vertical sequence into the BP neural network for flow rate prediction to obtain the second prediction result; perform mean processing on the first prediction result and the second prediction result to obtain the predicted flow rate of the air compressor in the future time period.

[0116] S5: Use the flow rate control method and the pressure control method to perform scheduling control on the air compressor unit, and determine whether the control errors of the flow rate control method and the pressure control method are within the first threshold range; if so, go to step S6; if not, go to step S7;

[0117] S6: Use the flow control method to schedule and control the air compressor unit, and output the corresponding control instructions;

[0118] S7: Use the pressure control method to schedule and control the air compressor unit, and output the corresponding control instructions.

[0119] In the present invention, the historical flow data of the air compressor is substituted into a hybrid prediction method combining horizontal prediction and vertical prediction to obtain the predicted air compressor flow result. Then, the original flow data, predicted flow data, and pressure data are substituted into the intelligent scheduling model to obtain the scheduling methods for different air compressors. Finally, the air compressor scheduling methods are sent to the main controller for execution to achieve intelligent scheduling. This method can be used for the control of the air compressor system to realize the intelligent scheduling of different types of air compressors and ultimately reduce the energy consumption of the air compressor unit.

[0120] In one example, the present invention also provides a storage medium, which has the same inventive concept as the intelligent scheduling method for the air compressor unit based on flow prediction formed by any one of the above examples or a combination of multiple examples. Computer instructions are stored thereon, and when the computer instructions run, they execute the steps of the intelligent scheduling method for the air compressor unit based on flow prediction formed by any one of the above examples or a combination of multiple examples.

[0121] Based on such an understanding, the technical solution of this embodiment, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks, etc., all of which can store program codes.

[0122] In one example, the present invention also provides a terminal, which has the same inventive concept as any one of the above examples or a combination of multiple examples corresponding to the intelligent scheduling method for the air compressor unit based on flow prediction. It includes a memory and a processor. Computer instructions that can run on the processor are stored on the memory, and when the processor runs the computer instructions, it executes the steps of the intelligent scheduling method for the air compressor unit based on flow prediction. The processor can be a single-core or multi-core central processing unit or a specific integrated circuit, or an integrated circuit configured to implement one or more of the present invention.

[0123] In one example, the terminal, i.e., the electronic device, is presented in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: the above-mentioned at least one processing unit (processor), the above-mentioned at least one storage unit, and a bus connecting different system components (including the storage unit and the processing unit).

[0124] Among them, the storage unit stores program code, and the program code can be executed by the processing unit, so that the processing unit executes the steps according to various exemplary embodiments of the present invention described in the "Exemplary Method" section of this specification. For example, the processing unit can execute the above-mentioned intelligent scheduling method of the air compressor unit based on traffic prediction.

[0125] The storage unit may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 3201 and / or a cache storage unit, and may further include a read-only storage unit (ROM).

[0126] The storage unit may also include a program / utility with a set (at least one) of program modules. Such program modules include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. The implementation of a network environment may be included in each or some combination of these examples.

[0127] The bus may represent one or more of several types of bus structures, including a storage unit bus or a storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any bus structure in a variety of bus structures.

[0128] The electronic device can also communicate with one or more external devices (such as a keyboard, a pointing device, a Bluetooth device, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device, and / or communicate with any device that enables the electronic device to communicate with one or more other computing devices (such as a router, a modem, etc.). Such communication can be carried out through an input / output (I / O) interface. Moreover, the electronic device can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter. The network adapter communicates with other modules of the electronic device through a bus. It should be understood that other hardware and / or software modules can be used in combination with the electronic device, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0129] Through the above description, those skilled in the art can easily understand that the exemplary embodiments described herein can be implemented by software or by a combination of software and necessary hardware. Therefore, the technical solution according to this exemplary embodiment can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method of the exemplary embodiment of the present application.

[0130] The above specific embodiments are detailed descriptions of the present invention. It cannot be determined that the specific embodiments of the present invention are only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions and substitutions can still be made, which should all be regarded as belonging to the protection scope of the present invention.

Claims

1. An intelligent scheduling method for air compressor units based on flow prediction, characterized in that: It includes the following steps: Judge whether the control errors of the flow control method and the pressure control method are within the first threshold range. If so, use the flow control method to perform scheduling control on the air compressor unit; If not, use the pressure control method to perform scheduling control on the air compressor unit; The flow control method includes: Predict the air compressor flow rate to obtain the predicted flow rate in the future time period; Determine the working state of the industrial frequency machine combination and the load rate of the frequency converter; Adjust the load rate of the frequency converter according to the predicted flow rate, so as to achieve the scheduling control of the air compressor unit; The pressure control method includes: Obtain the pressure value in the pipeline and judge whether the pressure value exceeds the second threshold; If it exceeds the lower limit of the second threshold, judge whether the frequency converter is fully loaded. If so, start at least one industrial frequency machine according to the full load duration of the frequency converter; If it exceeds the upper limit of the second threshold, judge whether the frequency converter is unloaded. If so, turn off at least one industrial frequency machine according to the unload duration of the frequency converter; The specific prediction of the air compressor flow rate includes: Divide the original air compressor flow data sequence into a horizontal sequence and a vertical sequence: the original air compressor flow data sequence is the horizontal sequence, extract the data in the original air compressor flow data sequence at intervals to obtain the horizontal sequence, and continue to extract the data in the horizontal sequence and / or the original air compressor flow data sequence at intervals to obtain more horizontal sequences; the relationship between the horizontal sequence and the vertical sequence is: x j = x i p ij ; where x j represents the data of the previous horizontal sequence; x i is the data of the next horizontal sequence; p ij is the vertical sequence. Input the horizontal sequence into the first neural network model for flow rate prediction to obtain the first prediction result; input the vertical sequence into the second neural network model for flow rate prediction to obtain the second prediction result; Perform mean processing or weight processing on the first prediction result and the second prediction result to obtain the predicted flow rate of the air compressor in the future time period.

2. The intelligent scheduling method for air compressor units based on flow prediction according to claim 1, wherein: After determining the working state of the industrial frequency machine combination and the load rate of the frequency converter, it further includes: Introduce the current pressure value and determine the working state of the industrial frequency machine combination and the load rate of the frequency converter again.

3. The intelligent scheduling method for air compressor units based on flow prediction according to claim 1, wherein: When adjusting the load rate of the frequency converter according to the predicted flow rate, make the load rate of the frequency converter be 60% - 80%.

4. The intelligent scheduling method for air compressor units based on flow prediction according to claim 1, characterized in that: When inputting the horizontal sequence into the first neural network model for flow rate prediction, it further includes: Perform stationarity test on the horizontal sequence.

5. The intelligent scheduling method for air compressor units based on flow prediction according to claim 1, characterized in that: Before dividing the original air compressor flow data into a horizontal sequence and a vertical sequence, it further includes: Perform data compensation processing on the original air compressor flow data sequence: when there are short-term data missing in the original air compressor flow data sequence, compensate according to the mean value of the adjacent data of the missing data; When there are long-term data missing in the original air compressor flow data sequence, compensate according to the historical data similarity.

6. The intelligent scheduling method for air compressor units based on flow prediction according to claim 1, characterized in that: Before inputting the horizontal sequence or the vertical sequence into the neural network model, it further includes: Perform Kalman filtering processing on the horizontal sequence and / or the vertical sequence.

7. A storage medium having computer instructions stored thereon, characterized in that: When the computer instruction runs, it executes the steps of the intelligent scheduling method of the air compressor unit based on flow rate prediction according to any one of claims 1-6.

8. A terminal, comprising a memory and a processor, where computer instructions that can run on the processor are stored on the memory, and characterized in that: When the processor runs the computer instruction, it executes the steps of the intelligent scheduling method of the air compressor unit based on flow rate prediction according to any one of claims 1-6.

Images