A diagnostic method and implementation system for internal faults of control system board channels

Abstract

The present invention discloses a method for diagnosing a control system board channel by applying an external signal and its implementation system. The diagnostic method comprises: connecting an external excitation signal source to the board process channel, starting the OPC Client, reading the process channel conversion value of the central control system server, obtaining the error time series formed by the signal source time series and the control system channel conversion time series, repeating this process multiple times to form a test sample, obtaining a channel model based on the sample data, performing a significance check on the mean and variance, and when both checks are accepted, the channel is diagnosed as being in a fault-free state. The present invention uses a data-driven method to detect the health status of the board internal channel, does not require knowledge of the mechanism of the board channel, does not require the addition of hardware equipment, and does not require the board to be disassembled, thereby reducing the professional knowledge requirements of the staff and lowering the diagnosis cost.

Description

Technical Field

[0001] The present invention belongs to the field of fault monitoring within control systems, and specifically relates to a method for diagnosing the health status of internal channels within control system boards. This method can be used to diagnose abnormal conditions or faults within control system boards, particularly when the internal channels cannot be directly observed. Background Art

[0002] Control systems are widely used across various sectors of the national economy. Failure in these systems can lead to decreased product quality, interrupted process flows, significant losses, and even safety incidents resulting in casualties. Control systems typically connect to field equipment through card channels. These card channels correspond to the actual device signal requirements and are generally categorized as analog input, analog output, digital input, and digital output. To ensure interference-free operation, control system cards are encapsulated within a housing. However, this encapsulation makes it impossible to measure the status of the card's internal channels, and forcibly damaging the encapsulation can cause unstable operation. Control systems typically provide fault codes to indicate internal channel failures, but the information these codes provide is very limited and cannot meet practical application requirements. Currently, diagnosis of control system cards requires return to the factory for inspection, but this presents challenges such as long maintenance cycles and high costs. Therefore, a method for detecting the health of card internal channels without disassembly has been developed, which has significant application value for on-site maintenance of control systems. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to apply an external signal to the channel to be detected, use OPC technology to obtain the conversion value of the control system to the process channel, and diagnose whether there is an abnormal condition in the channel by analyzing the error time series.

[0004] The technical solution adopted in the present invention is:

[0005] Step 1: Connect the external excitation signal source to the board process channel, set the detection time length to T, and obtain the signal source time series , k is the sampling time;

[0006] Step 2: Start the OPC Server of the central control system server, start the OPC Client, read the process channel conversion value of the central control system server, and record it as the control system channel conversion time series Here, OPC Server refers to the module within the central control system that can output data to third-party software. OPC Client refers to the interface program developed to access the control system's internal data, corresponding to the central control system's OPC Server.

[0007] Step 3: Use the formula , find the signal source time series and control system channel conversion time series The resulting error time series ,Right now

[0008] (1)

[0009] Where, is the sampling time, , n is the number of samples in the time series;

[0010] Step 4: Repeat steps 1 to 3 m times, with each detection time being T1, T2, ..., Tm, to obtain m sets of error time series , where is the number of trials, k is the number of samples for each test. , its mean and variance for

[0011] (2)

[0012] (3)

[0013] Step 5: Use formula (4) (5) to obtain the channel normal distribution model and its parameters

[0014] (4)

[0015] (5)

[0016] Step 6: Use the data acquisition and processing unit to obtain the error time series of the channel to be detected , and calculate the mean value through formula (2) (3) and variance , perform significance check:

[0017] 1. Mean significance test

[0018] Step 1: Assume H0: ; Calculate statistics ;

[0019] Step 2: Given a significance level , check the upper quantile table of the standard normal distribution to obtain ;

[0020] Step 3: If , then reject H0, if , then accept H0;

[0021] here is the mean of the test sample, is the mean of the channel normal model, is the variance of the test sample, is the number of trials to construct the channel normal model, is the upper quantile of the standard normal distribution;

[0022] 2. Variance significance test

[0023] Step 1: Assume H0: ; Statistics ;

[0024] Step 2: Given a significance level ,check The upper quantile table of the distribution is obtained and ;

[0025] Step 3: If or , then reject H0; if , then accept H0;

[0026] here is the variance of the test sample, is the variance of the channel normal model, is the number of trials to construct the channel normal model, and for The upper quantile of the distribution is obtained by looking up the table;

[0027] Step 7: When both checks are accepted, the channel is diagnosed as a no-fault state, otherwise it is diagnosed as a fault state.

[0028] Furthermore, the steps of the OPC Client are as follows:

[0029] Step 1: Get a list of servers

[0030] Step 2: Connect to the server,

[0031] Step 3: Expand all nodes in the server

[0032] Step 4: Select the node;

[0033] Step 5: Perform a read operation

[0034] Step 6: Write to the target database;

[0035] Step 7: Repeat steps 5 and 6 until finished.

[0036] The method of applying an external signal to diagnose a control system board channel of the present invention has the following beneficial effects compared with the prior art:

[0037] 1. This method is suitable for on-site control system board channel diagnosis. It can detect the health status of the board's internal channels without disassembling the board, so that effective measures can be taken in a timely manner to correct them, thus avoiding potential fault hazards.

[0038] 2. This method does not require knowledge of the mechanism of the board channel, and directly uses the external excitation signal source and the control system to convert data for analysis, thus greatly reducing the professional knowledge requirements of the staff;

[0039] This method utilizes the channel conversion data provided by the control system and can realize the diagnosis of the control system board channel without adding hardware equipment, thereby greatly reducing the cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 1 is a schematic structural diagram of a method for diagnosing a control system board channel by applying an external signal;

[0041] Figure 2 This is the OPC Client flow chart of the present invention. DETAILED DESCRIPTION

[0042] The following describes in detail the method for diagnosing a control system board channel using an external excitation signal according to the present invention in conjunction with the embodiments and drawings.

[0043] like Figure 1As shown, the present invention discloses a system for diagnosing the internal channels of control system boards by applying external signals. The system is used to diagnose the internal channels of control system boards and comprises: a central control system including a board to be tested and a server; an external excitation signal source for applying external signals to the channels to be tested; an OPC Client unit for acquiring converted data of the board channel stored in the server within the central control system; a data acquisition and processing unit for receiving a signal x from the external excitation signal source and data y from the converted board channel completed by the server and calculating the error z between the two sets of signals; a probability model unit for establishing a system probability model; and a channel diagnosis and conclusion output unit for fault diagnosis. The OPC Client refers to an interface program developed to acquire data within the control system corresponding to the OPC Server provided by the server of the central control system. The OPC Server refers to a module provided by the control system itself that can provide data output to third-party software. Third-party software refers to software not included in the control system product and application software compiled for service providers using the control system product as a platform.

[0044] Steps

[0045] The technical solution of the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. The specific embodiments described are only used to explain the present invention and are not intended to limit the present invention.

[0046] Step 1: Connect the signal generator to the process channel of the board, set the detection time length to T, and adjust the signal generator to emit a sine wave , in, is the sine wave amplitude, is the angular frequency of the sine wave. The sine signal is sampled within this time period to obtain the signal source time series , k is the sampling time;

[0047] Step 2: Start the OPC Server of the central control system server, start the OPC Client, read the process channel conversion value of the central control system server, and record it as the control system channel conversion time series Here, OPC Server refers to the module within the central control system that can output data to third-party software. OPC Client refers to the interface program developed to access the control system's internal data, corresponding to the central control system's OPC Server.

[0048] Step 3: Use the formula , find the signal source time series and control system channel conversion time series The resulting error time series ,Right now

[0049] (1)

[0050] Where, is the sampling time, , n is the number of samples in the time series;

[0051] Step 4: Repeat steps 1 to 3 m times, with each detection time being T1, T2, ..., Tm, to obtain m sets of error time series , where is the number of trials, k is the number of samples for each test. , its mean and variance for

[0052] (2)

[0053] (3)

[0054] Step 5: Use formula (4) (5) to obtain the channel normal distribution model and its parameters

[0055] (4)

[0056] (5)

[0057] Step 6: Use the data acquisition and processing unit to obtain the error time series of the channel to be detected , calculate the mean value through formula (2) (3) and variance , perform significance check:

[0058] 3. Mean significance test

[0059] Step 1: Assume H0: ; Calculate statistics ;

[0060] Step 2: Given a significance level , check the upper quantile table of the standard normal distribution to obtain ;

[0061] Step 3: If , then reject H0, if , then accept H0;

[0062] here is the mean of the test sample, is the mean of the channel normal model, is the variance of the test sample, is the number of trials to construct the channel normal model, is the upper quantile of the standard normal distribution;

[0063] 4. Variance significance test

[0064] Step 1: Assume H0: ; Statistics ;

[0065] Step 2: Given a significance level ,check The upper quantile table of the distribution is obtained and ;

[0066] Step 3: If or , then reject H0; if , then accept H0;

[0067] here is the variance of the test sample, is the variance of the channel normal model, is the number of trials to construct the channel normal model, and for The upper quantile of the distribution is obtained by looking up the table;

[0068] Step 7: When both checks are accepted, the channel is diagnosed as a no-fault state, otherwise it is diagnosed as a fault state.

[0069] The steps of the OPC client in this embodiment are as follows:

[0070] Step 1: Get a list of servers

[0071] Step 2: Connect to the server.

[0072] Step 3: Expand all nodes in the server

[0073] Step 4: Select a node;

[0074] Step 5: Perform a read operation

[0075] Step 6: Write to the target database;

[0076] Step 7: Repeat steps 5 and 6 until finished.

[0077] The steps for establishing the target database are as follows:

[0078] Step 1: Create an instance of SqlConnection, where SqlConnection is the database connection instruction provided by the operating system;

[0079] Step 2: Open the database connection.

[0080] The following experimental verification

[0081] The AI ​​channel was selected as the experimental object, and the data was acquired using an OPC client program written in VS2015. The excitation signal was implemented using a signal generator built in the central control system. The health data obtained 9 times was used as the observation sample set. The expectation and variance were calculated according to formulas (2) and (3). The results are shown in Table 1.

[0082] Table 1: Mean and variance of data samples (significance level 0.05)

[0083]

[0084] Based on the above sample parameters, using formulas (4) and (5), the mean of the channel normal distribution model obtained is -0.2050 and the variance is 0.3173;

[0085] Without destroying the board packaging, eight scenarios were simulated and tested by adding signals. These eight scenarios are shown in Table 2.

[0086] Table 2: 8 test conditions

[0087] Serial number Annotation state １ Scenario 1 Additional signal is 0, normal state is 1 ２ Scenario 2 Additional signal is 0, normal state 2 ３ Scenario 3 Additional signal is 0, normal state 3 ４ Scenario 4 Additional signal is 0, normal state is 4 ５ Scenario 5 Additional set value signal ６ Scenario 6 Additional sinusoidal wave signal ７ Scenario 7 Additional linear increase signal ８ Scenario 8 Add random noise signal

[0088] 200,000 signal source time series are randomly selected for each case and control system channel conversion time series The continuous records of are used as the test set. First, the error time series is calculated according to formula (2): , calculate the mean of the error time series and variance , and then the channel normal distribution model formed by the 9 groups of sub-samples was analyzed at the significant level. The results of the mean significance test and variance significance test are shown in Table 3;

[0089] Table 3: Channel diagnosis results (significance level 0.05)

[0090]

[0091] As can be seen from Table 3, through the significance check, cases 1-4 are concluded to be healthy, and cases 5-8 are concluded to be faulty, which is consistent with our previous assumptions.

Claims

1. A method for diagnosing a control system board channel by applying an external signal, characterized in that: The board channel diagnosis method comprises: Step 1: Connect the external excitation signal source to the board process channel, set the detection time length to T, and obtain the signal source time series {x(k)}, where k is the sampling time; Step 2: Start the OPC Server of the central control system server, start the OPC Client, read the process channel conversion value of the central control system server, and record it as the control system channel conversion time series {y(k)}; Step 3: Use the formula z = yx to calculate the error time series {z(k)} formed by the signal source time series {x(k)} and the control system channel conversion time series {y(k)}, where k = 1, 2, ..., n, where n is the number of samples in the time series; Step 4: Repeat steps 1 to 3 m times, with each test time being T1, T2, …, Tm, to obtain m sets of error time series {z(1, k)}, {z(2, k)}, …, {z(j, k)}, …, {z(m, k)}, where j = 1, …, m is the number of trials, and k = 1, …, n j is the number of samples sampled for each test; for the jth group of error time series {z(j,k)}, the mean μ(j) and variance σ are calculated according to the following formulas: 2 (j) Step 5: Use the following formula to obtain the channel normal distribution model N(μ0,σ0) and its parameters Step 6: Use the data acquisition and processing unit to obtain the error time series {z(k)} of the channel to be tested, calculate its mean μ and variance σ using the formula in step 4, and perform significance check according to the following method: (1) Mean significance check Step 1: Assume H0: μ = μ0; calculate the statistic Step 2: Given a significance level α, look up the upper quantile table of the standard normal distribution to obtain Step 3: If Then reject H0, if Then accept H0; Here μ is the mean of the test sample, μ0 is the mean of the channel normal model, σ is the variance of the test sample, and m is the number of trials for constructing the channel normal model. is the upper quantile of the standard normal distribution; (2) Variance significance test Step 1: Assume H0:σ=σ0; statistic Step 2: Given a significance level α, check χ 2 The upper quantile table of the distribution is obtained and Step 3: If or Then reject H0; if Then accept H0; Here σ is the variance of the test sample, σ0 is the variance of the channel normal model, and m is the number of trials for constructing the channel normal model. and is χ 2 The upper quantile of the distribution is obtained by looking up the table; Step 7: When both checks are accepted, the channel is diagnosed as a non-fault state, otherwise it is diagnosed as a fault state.

2. The method for diagnosing a control system board channel by applying an external signal according to claim 1, characterized in that: The detection time in step 1 is a finite time length T.

3. The method for diagnosing a system board channel by applying an external signal according to claim 1, characterized in that: The OPC Client described in step 2 includes: Get the server list of the central control system, connect to the server, expand all nodes in the server, select the node according to the board channel of the control system, perform the read operation, and write to the storage database.

4. The method for diagnosing a control system board channel by applying an external signal according to claim 1, characterized in that: The judgment of channel diagnosis is based on two significance checks: mean and variance.

5. A system for diagnosing a control system board channel by applying an external signal, wherein the system implements any one of claims 1 to 4, comprising: Central control system, OPC Client unit, data acquisition and processing unit, probability model unit, channel diagnosis and conclusion output unit; its characteristics are: The central control system includes DCS and PLC systems, including panels or modules for input and output channels; The OPC Client unit includes a software interface program or hardware device that is connected to the central control system server to obtain internal data of the control system; The data acquisition and processing unit includes a software program and hardware device for acquiring external excitation signals and channel conversion data; The probability model unit includes a normal distribution model after obtaining the mean and variance; The channel diagnosis and conclusion output unit includes mean significance check, variance significance check and output display.

Images