Data collection method and device capable of effectively improving data collection and transmission efficiency

Abstract

The present invention discloses a data acquisition method and device that can effectively improve the efficiency of data acquisition and transmission. The data acquisition method includes the following steps: setting an initialization threshold μ0, initial sampling {A0}, and initializing a warning value function; the system starts to perform low-frequency sampling to obtain multiple groups of low-frequency sampling data; using the least squares method to perform curve fitting on the multiple groups of low-frequency sampling data collected by the device, and obtains a fitting function, and uses the fitting function to calculate the data valuation at multiple subsequent time points; calculates the warning value function to determine the size relationship between the data valuation and the warning value; if the data valuation is exceeded, the system automatically switches to intermediate frequency sampling, and updates the fitting function through the intermediate frequency sampling data; uses the updated fitting function to calculate the data valuation at multiple subsequent time points; determines the size relationship between the data valuation and the threshold μ0; if the data valuation is exceeded, the system automatically switches to high frequency sampling.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation and intelligent manufacturing, and in particular to a data acquisition method and device that can effectively improve data acquisition and transmission efficiency. Background Art

[0002] Equipment data acquisition is the foundation for realizing the Industrial Internet and intelligent manufacturing. Conventional equipment data acquisition typically utilizes the existing network interface, RS232, or RS485 interface of the device's controller for direct data acquisition. When encountering different data protocols, an external data conversion module is often used for protocol conversion. When encountering different interface formats, an external data conversion module can also be used for interface conversion. Existing data acquisition methods mostly use direct acquisition, with external data conversion modules solely responsible for data format conversion or interface conversion. Some external data conversion modules also have data transmission capabilities, but most do not have data preprocessing capabilities. In terms of data transmission / output, most external data conversion modules use an acquisition-and-transmission approach.

[0003] The existing collection and transmission methods have the problem of large amounts of data collected and transmitted. When the equipment is operating normally, most of the operating data is stable or fluctuates in a normal period, and the data changes little or the fluctuation frequency is stable. Only when the equipment is operating abnormally or the production conditions change, the equipment operation data will send fluctuations or abnormal changes. These fluctuating or abnormally changing data are the important equipment operation data that data collection needs to capture. When using existing methods to collect equipment data, in order to capture those important equipment operation change data, shorter sampling intervals and higher data sampling frequencies are often used. However, higher data sampling frequencies will result in a very large amount of collected data and data transmission, which will greatly occupy data collection equipment and network resources, causing network congestion or increased network load, resulting in excessively high traffic costs.

[0004] Therefore, there is an urgent need to develop a method that can avoid collecting and transmitting a large amount of non-important data as much as possible when collecting equipment operation data, while capturing and collecting valid data from massive equipment operation data.

[0005] The information disclosed in this background technology section is only intended to enhance understanding of the overall background of the invention and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the Invention

[0006] The object of the present invention is to provide a data acquisition method and device that can effectively improve data acquisition and transmission efficiency, which can effectively improve data acquisition and transmission efficiency.

[0007] To achieve the above object, the present invention provides a data acquisition method that can effectively improve data acquisition and transmission efficiency, comprising the following steps:

[0008] Set the initial threshold μ0, the initial sampling {A0}, and initialize the warning value function;

[0009] The system starts to perform low-frequency sampling and obtains multiple sets of low-frequency sampling data;

[0010] The least squares method is used to perform curve fitting on multiple sets of low-frequency sampling data collected by the equipment, and the fitting function is obtained. Among them, a maximum error value is calculated after each fitting

[0011] Using the fitting function To calculate the data valuation at multiple subsequent time points

[0012] Calculate the warning value function The calculation method of the warning value is the threshold μ0 minus the maximum error value in the previous curve fitting;

[0013] Judging data valuation and warning value The size relationship;

[0014] If the data valuation The system automatically switches to intermediate frequency sampling and updates the fitting function through the intermediate frequency sampling data.

[0015] Using the updated fitting function To calculate the data valuation at multiple subsequent time points

[0016] Judging data valuation Relationship with the threshold μ0;

[0017] If the data valuation The system automatically switches to high-frequency sampling.

[0018] In a preferred embodiment, based on low-frequency sampling, if the estimated The system then performs the low-frequency sampling process again.

[0019] In a preferred embodiment, the intermediate frequency sampling time is set to last for one or two low frequency sampling cycles. If data exceeding the threshold is found during the intermediate frequency sampling cycle, the system automatically switches to high frequency sampling. High frequency sampling also lasts for one or two low frequency sampling cycles until the data tends to be stable. The system determines the subsequent data valuation. The system automatically switches to low-frequency sampling.

[0020] In a preferred embodiment, the intermediate frequency sampling frequency is set to 10 times the low frequency sampling frequency, and the high frequency sampling frequency is set to 10 times the intermediate frequency sampling frequency.

[0021] In a preferred embodiment, the initial sample {A0} is the actual collected data. Before executing the low-frequency sampling, the warning value is initialized according to the following formula:

[0022] In a preferred embodiment, the system records the intermediate frequency sampling data of the intermediate frequency sampling period and whether data exceeding the threshold μ0 appears in the intermediate frequency sampling period in real time, as an evolutionary scheme for system machine learning.

[0023] The present invention also discloses a data acquisition device that can effectively improve data acquisition and transmission efficiency, comprising:

[0024] A unit for initializing thresholds and warning values ​​and performing initial sampling, wherein the initialization threshold μ0 and the initial sampling {A0} are set;

[0025] A unit for performing low-frequency sampling to obtain multiple sets of low-frequency sampling data;

[0026] Used to perform curve fitting on multiple sets of low-frequency sampling data and obtain the fitting function The unit, in which the least square method is used for curve fitting, and a maximum error value is calculated after each fitting

[0027] Used to use the fitting function To calculate the data valuation at multiple subsequent time points Unit;

[0028] Used to calculate the warning value function The unit, where the warning value function The calculation method of is the threshold μ0 minus the maximum error value in the previous curve fitting;

[0029] Used to judge data valuation and warning value The unit of size relationship;

[0030] For data valuation The system automatically switches to intermediate frequency sampling and updates the fitting function through the intermediate frequency sampling data. Unit;

[0031] Fitting function for utilizing updates To calculate the data valuation at multiple subsequent time points Unit;

[0032] Used to judge data valuation The unit of the magnitude relationship with the threshold μ0;

[0033] For data valuation The system automatically switches to a high-frequency sampling unit; and

[0034] A unit for recording the intermediate frequency sampling data in the intermediate frequency sampling period and whether data exceeding the threshold μ0 appears in the intermediate frequency sampling period.

[0035] In a preferred embodiment, the intermediate frequency sampling time is set to last for one or two low frequency sampling cycles. If data exceeding the threshold is found during the intermediate frequency sampling cycle, the system automatically switches to high frequency sampling, and the high frequency sampling also lasts for one or two low frequency sampling cycles. The data acquisition device also includes a device for determining the subsequent data valuation when the high frequency sampling data tends to be stable. The system automatically switches to a low-frequency sampling unit.

[0036] Compared with the prior art, the data acquisition method and device of the present invention, which can effectively improve data acquisition and transmission efficiency, have the following beneficial effects: Traditional equipment data collection methods mostly use direct data collection, and the data collected by such methods contains a large amount of redundant data, which is not conducive to data processing and transmission. The data acquisition method of the present invention uses the mathematical least squares method to perform curve fitting, and uses the fitted curve to determine the possible trend of data changes in the next few cycles. This use of the fitted curve to predict subsequent sampled data, and by increasing the intermediate warning value and automatically and intelligently adjusting the data sampling frequency of production equipment multiple times, it avoids the collection of massive data during the data collection process while effectively capturing important fluctuations and changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a basic calculation logic diagram for adjusting the sampling frequency using the least squares method in the data acquisition application of the present invention.

[0038] Figure 2 Flowchart of the data collection method according to the preferred embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings. The embodiments of the present invention and all other embodiments obtained by persons of ordinary skill in the art without creative work are within the scope of protection of the present invention.

[0040] Unless expressly stated otherwise, throughout the specification and claims, the term "comprise" or variations such as "include" or "comprising", etc., will be understood to include the stated elements or components but not to exclude other elements or other components.

[0041] First, the principle of curve fitting using the least squares method of the present invention is described:

[0042] Let the unknown quantities to be determined be a1, a2, ..., a t , which can be obtained by n (n ≥ t) direct measurements y1, y2, …, y n , obtained through the following functional relationship:

[0043] y1=f1(a1,a2,…,a t )

[0044] y2=f2(a1,a2,…,a t )

[0045] y3=f3(a1,a2,…,a t )

[0046] ………

[0047] y n =f n (a1,a2,…,a t )

[0048] If a j is the true value, and the true value y is obtained from the above known function j , if its measured value is The corresponding error is (j=1,2,…n). The least squares method can be quantitatively expressed as:

[0049]

[0050] For measurements of different precision, the weight factor p of each measurement value should be added j ,Right now:

[0051]

[0052] The least squares method is a conclusion derived from the maximum likelihood method when the random error is normally distributed. It can minimize the sum of squares of measurement errors and is therefore considered the most reliable method to find a set of unknown quantities from a set of measurements.

[0053] The method of least squares polynomial curve fitting is as follows:

[0054] In order to solve the data fitting problem, the power function {1,x,x 2 ,…xn} as a function class, then

[0055]

[0056] This is the polynomial fit function.

[0057] To determine the fitting function The coefficients of , we need to solve the normal equations:

[0058]

[0059] It can also be expressed in matrix form as:

[0060]

[0061] The Vandermonde matrix can be simplified to:

[0062]

[0063] That is to say, X×A=Y, then A=(X×Y)-1×X×Y, and we get the coefficient matrix A, that is, we can solve a0, a1,…, a n , substitute it into The fitting polynomial can be obtained.

[0064] like Figure 1 As shown, after obtaining the fitting polynomial, by calculating That is, this polynomial can be used to calculate (predict) the data values ​​at multiple subsequent time points. And compare it with the pre-set data threshold value. If the predicted data will exceed the threshold μ0, the system starts high-frequency sampling and increases the collection points to obtain abnormal data or process change data of subsequent equipment operation. Otherwise, the system preset sampling frequency is still maintained to maintain a low amount of data and the amount of data transmission. This calculation method can accurately predict large-scale over-threshold phenomena, but it lacks the ability to predict high-frequency and low-amplitude over-threshold situations. At the same time, it is easy to produce errors in the prediction of data in a critical state. For example, long-term criticality may lead to continuous multiple high-frequency sampling. In this case, the demand for controlling the amount of data cannot be effectively met. Therefore, the present invention further optimizes this solution.

[0065] like Figure 2 As shown, the data acquisition method of the present invention, which can effectively improve the efficiency of data acquisition and transmission, includes the following steps:

[0066] Set the initial threshold μ0, the initial sampling {A0}, and initialize the warning value function;

[0067] The system starts to perform low-frequency sampling and obtains multiple sets of low-frequency sampling data;

[0068] The least squares method is used to perform curve fitting on multiple sets of low-frequency sampling data collected by the equipment, and the fitting function is obtained. Among them, a maximum error value is calculated after each fitting

[0069] Using the fitting function To calculate the data valuation at multiple subsequent time points

[0070] Calculate the warning value function The calculation method of the warning value is the threshold μ0 minus the maximum error value in the previous curve fitting;

[0071] Judging data valuation and warning value The size relationship;

[0072] If the data valuation The system automatically switches to intermediate frequency sampling and updates the fitting function through the intermediate frequency sampling data.

[0073] Using the updated fitting function To calculate the data valuation at multiple subsequent time points

[0074] Judging data valuation Relationship with the threshold μ0;

[0075] If the data valuation The system automatically switches to high-frequency sampling.

[0076] In a preferred embodiment, based on low-frequency sampling, if the estimated The system then performs the low-frequency sampling process again.

[0077] In a preferred embodiment, the intermediate frequency sampling time is set to last for one or two low frequency sampling cycles. If data exceeding the threshold is found during the intermediate frequency sampling cycle, the system automatically switches to high frequency sampling. High frequency sampling also lasts for one or two low frequency sampling cycles until the data tends to be stable. The system determines the subsequent data valuation. The system automatically switches to low-frequency sampling.

[0078] In a preferred embodiment, the intermediate frequency sampling frequency is set to 10 times the low frequency sampling frequency, and the high frequency sampling frequency is set to 10 times the intermediate frequency sampling frequency. The initialization threshold μ0 is set according to business requirements. The initial sampling {A0} is the actual collected data. Before executing low frequency sampling, the initial warning value is: This is the threshold.

[0079] In a preferred embodiment, the system records the intermediate frequency sampling data of the intermediate frequency sampling period and whether data exceeding the threshold μ0 appears in the intermediate frequency sampling period in real time, as an evolutionary scheme for system machine learning.

[0080] The data collection method of the present invention is suitable for industrial Internet data collection. Actual use results show that after improving the intelligent data collection technology algorithm, the data volume is reduced by 70% compared with the original data collection method, while still being able to collect more than 98% of super-threshold data. Therefore, it is an innovative method that can effectively improve data collection and transmission efficiency.

[0081] The present invention also discloses a data acquisition device that can effectively improve data acquisition and transmission efficiency, comprising:

[0082] A unit for initializing thresholds and warning values ​​and performing initial sampling, wherein the initialization threshold μ0 and the initial sampling {A0} are set;

[0083] A unit for performing low-frequency sampling to obtain multiple sets of low-frequency sampling data;

[0084] Used to perform curve fitting on multiple sets of low-frequency sampling data and obtain the fitting function The unit, in which the least square method is used for curve fitting, and a maximum error value is calculated after each fitting.

[0085] Used to use the fitting function To calculate the data valuation at multiple subsequent time points Unit;

[0086] Used to calculate the warning value function The unit, where the warning value function The calculation method of is the threshold μ0 minus the maximum error value in the previous curve fitting;

[0087] Used to judge data valuation and warning value The unit of size relationship;

[0088] For data valuation The system automatically switches to intermediate frequency sampling and updates the fitting function through the intermediate frequency sampling data. Unit;

[0089] Fitting function for utilizing updates To calculate the data valuation at multiple subsequent time points Unit;

[0090] Used to judge data valuation The unit of the magnitude relationship with the threshold μ0;

[0091] For data valuation The system automatically switches to a high-frequency sampling unit; and

[0092] A unit for recording the intermediate frequency sampling data in the intermediate frequency sampling period and whether data exceeding the threshold μ0 appears in the intermediate frequency sampling period.

[0093] In a preferred embodiment, the intermediate frequency sampling time is set to last for one or two low frequency sampling cycles. If data exceeding the threshold is found during the intermediate frequency sampling cycle, the system automatically switches to high frequency sampling, and the high frequency sampling also lasts for one or two low frequency sampling cycles. The data acquisition device also includes a device for determining the subsequent data valuation when the high frequency sampling data tends to be stable. The system automatically switches to a low-frequency sampling unit.

[0094] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A data acquisition method that can effectively improve data acquisition and transmission efficiency, characterized by: The steps include: Set the initial threshold μ0, the initial sampling {A0}, and initialize the warning value function; The system starts to perform low-frequency sampling and obtains multiple sets of low-frequency sampling data; The least squares method is used to perform curve fitting on multiple sets of low-frequency sampling data collected by the equipment, and the fitting function is obtained. Among them, a maximum error value is calculated after each fitting Using the fitting function To calculate the data valuation at multiple subsequent time points Calculate the warning value function The calculation method of the warning value is the threshold μ0 minus the maximum error value in the previous curve fitting; Judging data valuation and warning value The size relationship; If the data valuation ≤ warning value The system automatically switches to intermediate frequency sampling and updates the fitting function through the intermediate frequency sampling data. Using the updated fitting function To calculate the data valuation at multiple subsequent time points Judging data valuation Relationship with the threshold μ0; If the data valuation The system automatically switches to high-frequency sampling; The data acquisition method further comprises the following steps: based on low-frequency sampling, if the estimated >Alert value Then the system re-performs the low-frequency sampling process; The intermediate frequency sampling time is set to last for one or two low frequency sampling cycles. If data exceeding the threshold is found during the intermediate frequency sampling cycle, the system automatically switches to high frequency sampling. High frequency sampling also lasts for one or two low frequency sampling cycles until the data tends to be stable. The system then determines the subsequent data valuation. The system automatically switches to low-frequency sampling; The initial sampling {A0} is the actual collected data. Before executing low-frequency sampling, the warning value is initialized according to the following formula: φ* / i<0=μ0; The system records the intermediate frequency sampling data of the intermediate frequency sampling period in real time and whether data exceeding the threshold μ0 appears in the intermediate frequency sampling period as an evolutionary scheme for system machine learning.

2. The data collection method according to claim 1, wherein: The intermediate frequency sampling frequency is set to 10 times the low frequency sampling frequency, and the high frequency sampling frequency is set to 10 times the intermediate frequency sampling frequency.

3. A data acquisition device that can effectively improve data acquisition and transmission efficiency, characterized by: include: A unit for initializing thresholds and warning values ​​and performing initial sampling, wherein the initialization threshold μ0 and the initial sampling {A0} are set; A unit for performing low-frequency sampling to obtain multiple sets of low-frequency sampling data; Used to perform curve fitting on the multiple sets of low-frequency sampling data and obtain the fitting function The unit, in which the least square method is used for curve fitting, and a maximum error value is calculated after each fitting For using the fitting function To calculate the data valuation at multiple subsequent time points Unit; Used to calculate the warning value function The unit, where the warning value function The calculation method of is the threshold μ0 minus the maximum error value in the previous curve fitting; Used to judge data valuation and warning value The unit of size relationship; For data valuation ≤ warning value The system automatically switches to intermediate frequency sampling and updates the fitting function through the intermediate frequency sampling data. Unit; Fitting function for utilizing updates To calculate the data valuation at multiple subsequent time points Unit; Used to judge data valuation The unit of the magnitude relationship with the threshold μ0; For data valuation The system automatically switches to a high-frequency sampling unit; and A unit for recording intermediate frequency sampling data in an intermediate frequency sampling period and whether data exceeding a threshold μ0 appears in the intermediate frequency sampling period; Based on low-frequency sampling, if the estimation >Alert value Then the system re-performs the low-frequency sampling process; The intermediate frequency sampling time is set to last for one or two low frequency sampling cycles. If data exceeding the threshold is found during the intermediate frequency sampling cycle, the system automatically switches to high frequency sampling. High frequency sampling also lasts for one or two low frequency sampling cycles until the data tends to be stable. The system then determines the subsequent data valuation. The system automatically switches to low-frequency sampling.

Images