Remote control system and control method for power supply of injection molding machine

Abstract

The invention relates to a remote control system and control method for a power supply of an injection molding machine, and belongs to the technical field of power supply control, the remote control system comprises a control module, an execution module and a detection module which are electrically connected in sequence, and the detection module is used for obtaining power consumption data of power utilization changing along with time and uploading the power consumption data to the control module; the control module is used for archiving power consumption data according to time, predicting power consumption peak frequency x and power consumption y in a peak power consumption period by using a machine learning algorithm, and uploading time and power consumption to the execution module, the execution module turns on a standby power supply f (x) ahead of the peak power consumption period and increases the power by g (y) times, f (x) = kx + c, g (y) = jy + d, and the execution module is connected with the standby power supply. K, c, j and d are constants input to the control module in advance by an operator according to the redundancy; power supply work can be planned in advance, and the intelligent degree is high.

Description

Technical Field

[0001] The invention belongs to the technical field of power supply control, and in particular relates to a power supply remote control system and a control method for an injection molding machine. Background Art

[0002] In injection molding production, the injection molding machine is the main equipment. As a large equipment with high power consumption and multiple parts, such as injection molding motors, it is necessary to control the power supply of the injection molding machine.

[0003] In most cases, a general power remote control system can ensure stable and reliable operation, but occasionally there will be a phenomenon of application software or system crash and freeze. In order to complete the restart of the equipment system and application software, the existing operation method is that the staff goes downstairs to manually restart the equipment power supply. Such operation affects the recovery time of the line, wastes manpower, and reduces work efficiency. For this reason, the power remote control system disclosed in Chinese patent CN112510823B includes: a main control module, a controlled module, a power control box and a controlled device; the main control module, the controlled module, the power control box and the controlled device are connected in sequence; the main control module is used to detect the operating status of the controlled device in real time, and display the information of the controlled device in a graphical visual human-machine interface; the controlled module is used to receive the control command of the main control module, and convert the control command into an operation instruction that can be recognized by the power control box; the power control box includes a serial port controller and a DC relay, and the serial port controller controls the power on and off of the controlled device through the DC relay. It can detect and control the power on and off of the controlled device online, realize remote operation of the controlled device, save manpower, and improve efficiency;

[0004] However, the above scheme can only realize simple remote control of power supply and has basic control functions, such as wireless control of power supply without relying on a wired control panel. However, the injection molding machine consumes a large amount of electricity intermittently during use, and the power of the entire system fluctuates greatly over time, requiring advance planning of power supply operation. The above scheme cannot plan the power supply operation in advance, and the control intelligence is not high. Therefore, a power remote control system and control method for an injection molding machine that can plan the power supply operation in advance and has a high degree of intelligence is needed. Summary of the invention

[0005] In order to solve the above problems existing in the prior art, the present invention provides a power remote control system and control method for an injection molding machine, which has the characteristics of being able to plan the power work in advance and having a high degree of intelligence.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A power remote control system for an injection molding machine comprises a control module, an execution module and a detection module which are electrically connected in sequence, wherein the detection module is used to obtain power consumption data of power consumption varying with time and upload the data to the control module, the control module is used to archive the power consumption data by time and predict the peak power consumption frequency x and power consumption y during the peak power consumption period by using a machine learning algorithm, and upload the time and power consumption to the execution module, and the execution module turns on the backup power supply f(x) in advance during the peak power consumption period and increases the power g(y) times, wherein f(x)=kx+c, g(y)=jy+d, and k, c, j and d are all constants input into the control module in advance by an operator according to redundancy.

[0008] As a preferred technical solution of the present invention, the control module determines algorithm parameters (p, d, q) using the ARIMA algorithm based on the power consumption data archived by time, and uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period.

[0009] As a preferred technical solution of the present invention, the control module is pre-input with a working cycle time period, and the control module determines whether the peak power consumption frequency in the previous working cycle at the current time point exceeds a threshold. When the judgment result is yes, the control module increases the value of the parameter p and re-uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period.

[0010] As a preferred technical solution of the present invention, the control module is pre-input with a peak power frequency threshold u0, and the control module determines whether the peak power frequency u in the previous working cycle at the current time point exceeds the threshold u0. When the judgment result is yes, the control module corrects the parameter p to p×A1 and re-uses the ARIMA algorithm to predict the peak power frequency x and power consumption y during the peak power consumption period;

[0011] Where A1=u / u0×c, (p×A1)≤5.

[0012] As a preferred technical solution of the present invention, the control module corrects the parameter p to p×A1 and reuses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, and then calculates the mean of the calculation result and the peak power consumption frequency x and power consumption y predicted by the ARIMA algorithm using the parameter p during the peak power consumption period to calculate the corrected peak power consumption frequency x1 and the corrected power consumption y1. The execution module turns on the backup power supply f(x1) in advance during the peak power consumption period and increases the power g(y1) times.

[0013] As a preferred technical solution of the present invention, it also includes a fault detection module, which is used to determine whether the power consumption of the injection molding machine and the backup power supply exceeds a threshold value, and when the judgment result is yes, an alarm is issued through an alarm.

[0014] As a preferred technical solution of the present invention, it also includes a control panel, and the control panel is used to input the values ​​of k, c, j and d.

[0015] A method for remotely controlling the power supply of an injection molding machine is applicable to the remote control system of the power supply of the injection molding machine, and is characterized in that it comprises the following steps:

[0016] Step 1: The sensor collects the power output and uploads it to the control module in chronological order;

[0017] Step 2: The control module calculates the parameters (p, d, q) according to the output power values ​​received in time sequence, and uses the parameters (p, d, q) to establish the ARIMA model;

[0018] Step 3: The control module counts the peak power consumption frequency x and power consumption y during the peak power consumption period;

[0019] Step 4: The control module instructs the module to turn on the backup power supply f(x) in advance during the peak power consumption period and increase the power g(y) times, where f(x)=kx+c, g(y)=jy+d, k, c, j and d are all constants input into the control module in advance by the operator according to the redundancy.

[0020] The beneficial effects of the present invention are:

[0021] (1) By enabling the control module to archive power consumption data by time and using a machine learning algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, and instructing the execution module to turn on the backup power supply f(x)=kx+c in advance during the peak power consumption period and increase the power g(y)=jy+d times, the power supply work can be planned in advance, thereby improving the degree of intelligence;

[0022] (2) By making the control module determine whether the peak power consumption frequency in the previous working cycle at the current time point exceeds the threshold, and when the judgment result is yes, increasing the value of parameter p and reusing the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, it is ensured that when there are more representative sample data and less overfitting risk, the model complexity can be appropriately increased. When there are fewer sample data, a conservative model is used to reduce the risk of overfitting;

[0023] (3) After the control module reuses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, the calculation result is averaged with the ARIMA algorithm predicted by parameter p to obtain the corrected peak power consumption frequency x1 and corrected power consumption y1. When one data deviates, the degree of deviation of x1 or y1 is smaller than that of the deviation data, thereby reducing the risk caused by data deviation. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0025] Figure 1 is a control module block diagram of the present invention; DETAILED DESCRIPTION

[0026] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0027] See also Figure 1 , a power remote control system and control method for an injection molding machine, comprising a control module, an execution module and a detection module electrically connected in sequence, wherein the detection module is used to obtain power consumption data of power consumption varying with time and upload it to the control module;

[0028] Specifically, the detection module includes at least a power monitoring device, such as a smart meter, which collects the power data of the injection molding machine and the time of collecting the data in real time, and matches the power consumption data and time one by one, assigns time data to each power consumption data, and uploads it to the control module;

[0029] The control module includes at least a calculation and data storage module;

[0030] The execution module includes a backup power supply and a control circuit for controlling the backup power supply, the control circuit is used to receive instructions from the control module and instruct the backup power supply to operate according to the instructions, and the backup power supply is electrically connected to the injection molding machine and supplies power to the injection molding machine;

[0031] After receiving the data, the control module archives the power consumption data after assigning the time data by time, and then forms an image of power change according to time, which is recorded as historical data;

[0032] Optionally, the historical data may be preprocessed, including using statistical methods such as Z-Score to identify and remove outliers, and using moving average or wavelet transform methods to eliminate noise in the data;

[0033] Subsequently, the control module is used to archive the power consumption data by time and use a machine learning algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period;

[0034] Specifically, the control module uses a machine learning algorithm to predict the peak power consumption period once every hour. After the prediction is made using the machine learning algorithm, an image of the power consumption change over time in the next 24 hours is obtained based on the current trend. Then the control module determines whether the number of peak values ​​within 12 consecutive hours exceeds a threshold. When the judgment result is yes, the control module determines that this time period is a peak power consumption period. When the number of peak values ​​within several 12-hour time periods exceeds the threshold, the 12-hour time period with the largest number of peak values ​​is taken as the peak power consumption period.

[0035] Then, the control module counts the peak power consumption frequency x and power consumption y;

[0036] Specifically, the peak power consumption frequency x is the number of peaks in the graph of power consumption changing over time, and the power consumption y is the average value of several peak power consumption values;

[0037] At this time, the peak power consumption frequency x represents the frequency of the injection molding machine operation during the future peak power consumption period, and the power consumption y represents the average value of the peak power consumption of the injection molding machine;

[0038] Therefore, the control module instructs the execution module to turn on the backup power supply f(x) in advance during the peak power consumption period and increase the power g(y) times, where f(x) = kx + c, g(y) = jy + d, and k, c, j, and d are all constants input into the control module in advance by the operator according to the redundancy;

[0039] When x is larger, it means that the injection molding machine runs more frequently during the peak period of electricity consumption, and the probability of frequent operation before the peak period is higher. The backup power system needs to be turned on in advance to supply power to the injection molding machine. At this time, the degree of advance operation needs to be increased. At this time, the value of f(x) is larger, and the backup system is instructed to increase the degree of advance operation when x is larger.

[0040] When x is smaller, it means that the probability of the injection molding machine starting to run frequently before the peak period is smaller. Starting the backup power supply too early will lead to energy consumption. There is no need to turn on the backup power supply too much in advance. At this time, the value of f(x) is smaller, and the backup system is instructed to reduce the degree of early operation when x is smaller.

[0041] When y is larger, it means that more electricity is consumed during the peak period, and the backup power supply needs to start the backup with a higher power. At this time, the value of g(y)=jy+d is larger, and the backup power supply is instructed to increase the operating power when more electricity is consumed during the peak period.

[0042] When y is smaller, it means that less power is consumed during the peak hours. Starting the backup power supply at too high a power will lead to energy consumption, and the backup power supply needs to be started at a higher power. At this time, the value of g(y)=jy+d is smaller, and the backup power supply is instructed to reduce the operating power when less power is consumed during the peak hours.

[0043] At the same time, in actual use, the operator can adjust the values ​​of k, c, j and d according to the redundancy requirements. When a large redundancy is required for the advance start time, the operator increases the input k and c values. At this time, the value of f(x) is larger when the x value is the same, leaving a large redundancy for the advance start time. When a large redundancy is not required, the operator reduces the input k and c values.

[0044] Similarly, when a large margin is required for power, the operator increases the input values ​​of j and d, and when a large margin is not required, the operator decreases the input values ​​of j and d;

[0045] By enabling the control module to archive power consumption data by time and using a machine learning algorithm to predict peak power consumption frequency x and power consumption y during peak power consumption periods, and instructing the execution module to turn on the backup power supply f(x)=kx+c in advance during the peak power consumption period and increase the power g(y)=jy+d times, the power supply work can be planned in advance, thereby improving the degree of intelligence.

[0046] In the process of the control module using the machine learning algorithm to predict the peak power consumption frequency x and the power consumption y during the peak power consumption period, specifically, the control module uses the ARIMA algorithm to determine the algorithm parameters (p, d, q) according to the power consumption data archived by time, and uses the ARIMA algorithm to predict the peak power consumption frequency x and the power consumption y during the peak power consumption period;

[0047] After the control module receives the historical data and pre-processes it, each time a forecast is made, the unit root test is first used to check the stationarity of the data. In this embodiment, the method is the ADF test;

[0048] Subsequently, after checking the stability of the data, the control module determines whether the historical data is stable. If the judgment result is no, the historical data is differentially operated to achieve stability. Specifically, the original data is firstly differentially operated, and whether the historical data is stable is re-judged. If the judgment result is no, the differential operation is repeated until the data reaches a stable state. At this time, the required number of differential times d is recorded;

[0049] Then the control module calculates the autocorrelation function (ACF) and partial autocorrelation function (PACF) according to different p and q values, analyzes the ACF and PACF graphs under different p and q values, and selects the combination with the minimum information criterion. The p and q values ​​corresponding to this combination are the autoregressive order p and the sliding average order q.

[0050] At this time, the control module determines the algorithm parameters (p, d, q) and establishes the ARIMA (p, d, q) model. After the establishment, the established ARIMA model is used to obtain the image of the change of power consumption over time in the future period, and the image is used to count the peak power consumption frequency x and power consumption y during the peak power consumption period.

[0051] In the above process, according to the definition of the ARIMA (p, d, q) model, when the parameter p becomes larger, the complexity of the model will increase, and more complex patterns can be captured, but there is a probability of overfitting, especially when the sample data is small, the risk of overfitting is high. When the parameter p becomes smaller, the model becomes simpler, and there is a probability of losing the characteristics of the model, which reduces the accuracy of the forecast. Therefore, when there are more peak data in the past period of time, it means that there are more sample data of peak power consumption data, and the risk of overfitting is less when predicting peak power consumption. Therefore, appropriately increasing the value of parameter p increases the complexity of the model and improves the prediction accuracy. When there are fewer peak data, it means that there are fewer sample data of peak power consumption data, and the risk of overfitting is greater. At this time, it is necessary to appropriately reduce the value of parameter p to reduce the probability of overfitting and improve the prediction accuracy.

[0052] To this end, the control module is pre-input with a working cycle time period, and the control module determines whether the peak power consumption frequency in the previous working cycle at the current time point exceeds the threshold. When the judgment result is yes, the control module increases the value of the parameter p and re-uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period;

[0053] Specifically, a working cycle is twelve hours or is manually determined by the operator according to the production plan, and the last fixed period when the control module performs calculations every hour is a working cycle;

[0054] When the peak power consumption frequency in the previous working cycle exceeds the threshold, it means that the value of p can be increased to re-predict and obtain a more accurate prediction result. At this time, the control module increases the value of parameter p and re-uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, and instructs the execution module to use the recalculated x and y as the calculation parameters of f(x) and g(y);

[0055] Specifically, the control module is pre-input with a peak power frequency threshold u0. The control module determines whether the peak power frequency u in the previous working cycle at the current time point exceeds the threshold u0. When the judgment result is yes, the control module corrects the parameter p to p×A1 and re-uses the ARIMA algorithm to predict the peak power frequency x and power consumption y during the peak power consumption period; wherein A1=u / u0×c, (p×A1)≤5, c is a pre-input constant, and when p×A1 is greater than 5, the control module takes p×A1=5;

[0056] When u exceeds the threshold u0, it means that the value of p can be increased to re-predict and obtain a more accurate prediction result. At this time, the value of A1=u / u0×c is large. When the control module corrects the parameter p to p×A1 and reuses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, it is completed to increase the value of parameter p when there are more sample data and reuse the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period;

[0057] By making the control module determine whether the peak frequency of electricity consumption in the previous working cycle at the current time point exceeds the threshold, and when the judgment result is yes, increasing the value of parameter p and reusing the ARIMA algorithm to predict the peak frequency x and power consumption y during the peak electricity consumption period, it is ensured that when there are more representative sample data and less overfitting risk, the complexity of the model can be appropriately increased. When there are fewer sample data, a conservative model is used to reduce the risk of overfitting.

[0058] When the peak power consumption frequency exceeds the threshold and is recalculated, although the risk of overfitting is small when the sample size is large, there is still a risk of overfitting, and the obtained values ​​of x and y have a probability of low prediction accuracy. In order to reduce this part of the risk, the control module corrects the parameter p to p×A1 and re-uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, and then calculates the mean of the calculation result and the peak power consumption frequency x and power consumption y predicted by the ARIMA algorithm using the parameter p during the peak power consumption period, and calculates the corrected peak power consumption frequency x1 and the corrected power consumption y1. The execution module turns on the backup power supply f(x1) in advance during the peak power consumption period and increases the power g(y1) times;

[0059] After the mean calculation, the values ​​of x1 and y1 are between the x and y values ​​obtained before and after the recalculation. When one data deviates, the degree of deviation of x1 or y1 is smaller than that of the deviated data, reducing the risk of data deviation.

[0060] After the control module reuses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period, the calculation results are averaged with the ARIMA algorithm predicted by parameter p to obtain the corrected peak power consumption frequency x1 and corrected power consumption y1. When one data deviates, the degree of deviation of x1 or y1 is smaller than that of the deviation data, thereby reducing the risk caused by data deviation.

[0061] In order to promptly report a fault to the operator when it occurs, a fault detection module is also included. The fault detection module is used to determine whether the power consumption of the injection molding machine and the backup power supply exceeds a threshold value. When the judgment result is yes, an alarm is sounded through an alarm.

[0062] In order to facilitate the input of constants such as k, c, j and d, a control panel is also included. The control panel is used to input constants such as k, c, j and d, and is also used to display a graph of power consumption changes over time.

[0063] The present invention also provides a power remote control method for an injection molding machine, which is applicable to the power remote control system of the injection molding machine described above, and comprises the following steps:

[0064] Step 1: The sensor collects the power output and uploads it to the control module in chronological order;

[0065] Step 2: The control module calculates the parameters (p, d, q) according to the output power values ​​received in time sequence, and uses the parameters (p, d, q) to establish the ARIMA model;

[0066] Step 3: The control module counts the peak power consumption frequency x and power consumption y during the peak power consumption period;

[0067] Step 4: The control module instructs the module to turn on the backup power supply f(x) in advance during the peak power consumption period and increase the power g(y) times, where f(x)=kx+c, g(y)=jy+d, k, c, j and d are all constants input into the control module in advance by the operator according to the redundancy.

[0068] The working principle and use process of the present invention:

[0069] The detection module obtains power consumption data that changes with time and uploads it to the control module. The control module is used to archive the power consumption data by time and use a machine learning algorithm to predict the peak frequency x and power consumption y during the peak power consumption period, and upload the time and power consumption to the execution module. The execution module turns on the backup power supply f(x) in advance during the peak power consumption period and increases the power g(y) times.

[0070] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A remote power control system for an injection molding machine, characterized in that: It includes a control module, an execution module and a detection module which are electrically connected in sequence. The detection module is used to obtain power consumption data of electricity consumption changing with time and upload it to the control module. The control module is used to archive the power consumption data by time and use a machine learning algorithm to predict the peak frequency x and power consumption y of electricity consumption during the peak power consumption period, and upload the time and power consumption to the execution module. The execution module turns on the backup power supply f(x) in advance during the peak power consumption period and increases the power g(y) times, wherein f(x)=kx+c, g(y)=jy+d, and k, c, j and d are all constants input into the control module in advance by the operator according to redundancy.

2. The power remote control system of an injection molding machine according to claim 1, characterized in that: The control module determines algorithm parameters (p, d, q) using the ARIMA algorithm according to the power consumption data archived by time, and predicts the peak power consumption frequency x and power consumption y during the peak power consumption period using the ARIMA algorithm.

3. The power remote control system of an injection molding machine according to claim 1, characterized in that: The control module is pre-input with a working cycle time period. The control module determines whether the peak power consumption frequency in the previous working cycle at the current time point exceeds a threshold. When the judgment result is yes, the control module increases the value of the parameter p and re-uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period.

4. The power remote control system for an injection molding machine according to claim 3, characterized in that: The control module is pre-input with a peak power frequency threshold u0. The control module determines whether the peak power frequency u in the previous working cycle at the current time point exceeds the threshold u0. When the judgment result is yes, the control module corrects the parameter p to p×A1 and re-uses the ARIMA algorithm to predict the peak power frequency x and power consumption y during the peak power consumption period; Where A1=u / u0×c, (p×A1)≤5.

5. The power remote control system for an injection molding machine according to claim 4, characterized in that: The control module corrects the parameter p to p×A1 and re-uses the ARIMA algorithm to predict the peak power consumption frequency x and power consumption y during the peak power consumption period. Then, the calculation result is averaged with the peak power consumption frequency x and power consumption y predicted by the ARIMA algorithm using the parameter p to calculate the corrected peak power consumption frequency x1 and the corrected power consumption y1. The execution module turns on the backup power supply f(x1) in advance during the peak power consumption period and increases the power g(y1) times.

6. The power remote control system for an injection molding machine according to claim 1, characterized in that: It also includes a fault detection module, which is used to determine whether the power consumption of the injection molding machine and the backup power supply exceeds a threshold value, and when the judgment result is yes, an alarm is issued through an alarm.

7. The remote control system for power supply of an injection molding machine according to claim 1, characterized in that: Also included is a control panel for inputting values ​​of k, c, j and d.

8. A method for remotely controlling the power supply of an injection molding machine, applicable to the remotely controlling system for the power supply of an injection molding machine according to any one of claims 2 to 7, characterized in that: The following steps are involved: Step 1: The sensor collects the power output and uploads it to the control module in chronological order; Step 2: The control module calculates the parameters (p, d, q) according to the output power values ​​received in time sequence, and uses the parameters (p, d, q) to establish the ARIMA model; Step 3: The control module counts the peak power consumption frequency x and power consumption y during the peak power consumption period; Step 4: The control module instructs the module to turn on the backup power supply f(x) in advance during the peak power consumption period and increase the power g(y) times, where f(x)=kx+c, g(y)=jy+d, k, c, j and d are all constants input into the control module in advance by the operator according to the redundancy.

Images