High-speed analog-to-digital conversion sampling stability improving method based on improved Wiener filtering

Abstract

The invention belongs to the technical field of railway locomotive part detection signal processing, and particularly relates to a high-speed analog-to-digital conversion sampling stability improving method based on improved Wiener filtering, which comprises the following steps of: acquiring an original analog signal, and converting to obtain a digital semaphore; collecting and measuring a field noise environment; wiener filtering superposition is carried out on the interference characteristics to generate a conversion digital quantity signal; in the A / D starting working process, setting a self-adaptive Wiener filtering selection identifier 0; selecting a noise identification condition according to self-adaption; and establishing a feedback mechanism. According to the method, improved Wiener filtering is introduced, so that the anti-noise capability and the sampling stability after analog quantity to digital quantity conversion are remarkably improved, and the limitation of a traditional method in processing a complex noise environment is effectively solved. The algorithm not only can maintain high fidelity of signals, but also can remarkably improve the precision of data conversion and the overall robustness of the system, and is suitable for multiple fields of high-precision measurement, signal processing, communication systems and the like.

Description

Technical Field

[0001] The present invention belongs to the technical field of railway locomotive vehicle component detection signal processing, and in particular relates to a method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering. Background Art

[0002] In the on-site environment of rail transit locomotive vehicle brake valve detection, the collected signal will be interfered by various noise sources. The usual practice is to use filtering to filter out the noise that interferes with our actual true value signal and separate the desired signal. Traditional Wiener filtering requires that the input process of the signal is stable and random, and Wiener filtering can be used whether it is continuous or discrete. However, its shortcomings are also obvious: the input signal and the interference noise signal are required to be stable and random and the spectral characteristics are known. However, in practical applications, due to environmental noise, circuit non-ideal factors and quantization errors, the characteristics of being interfered by noise are often unknown, and it is difficult to meet the filtering premise of Wiener filtering. Therefore, Wiener filtering cannot achieve its optimal filtering purpose and is greatly limited, resulting in fluctuations and distortions in the sampled data, which is difficult to meet the high-precision and high-stability sampling requirements in the subsequent test process, affecting the accuracy of subsequent data processing and analysis. Summary of the invention

[0003] In view of the technical problems mentioned above, during the inspection process of key components of railway vehicles, due to the selection of circuit system components and on-site electromagnetic noise, the analog signal output by the measurement sensor is superimposed and input into the digital-to-analog conversion chip of the computer high-speed acquisition control card, and the noise is synchronously converted after analog-to-digital conversion, which causes distortion of the actual detection data, resulting in poor sampling stability of the system software, and the sampling board is affected by the on-site environment and has low adaptability. The present invention provides a method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering. By introducing the improved Wiener filtering technology, accurate optimization of the sampled digital quantity is achieved, and the accuracy of data conversion and the overall stability of the system are improved.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering includes the following steps:

[0006] S1. Remove high-frequency noise and interference from the power supply system of the current sensor for component detection and the A / D transmission line to the acquisition card, obtain the original analog signal, ensure that the effective information within the signal frequency band is not lost, and convert it into a digital signal;

[0007] S2. Collect and measure the noise environment on site, and collect the characteristics of the switch power supply noise, cable connector noise, and the start-stop noise of surrounding high-power equipment on site;

[0008] S3, performing Wiener filtering and superposition on the interference characteristics to generate a converted digital signal;

[0009] S4, during the A / D startup process, set the adaptive Wiener filter selection flag 0;

[0010] S5, according to the adaptive selection noise recognition situation;

[0011] S6. Establish a feedback mechanism to compare the sampled digital signal with the original analog signal, calculate the error, and feed the error information back to the improved Wiener filter for iterative optimization until the preset accuracy requirement is achieved.

[0012] The method for converting the digital signal quantity in S1 is as follows: the algorithm defines the target signal as s(t), the A / D conversion chip precision is a bit, the corresponding conversion voltage maximum value is b, the fixed conversion ratio is k=2a / b, and the converted digital signal quantity is expressed as ds(t)=k*s(t).

[0013] The analog noise signal of the on-site noise environment in S2 is defined as m(t), n(t), o(t), and the digital noise signal is defined as dm(t), dn(t), do(t).

[0014] The converted digital signals in S3 are d1(t)=ds(t)+dm(t), d2(t)=ds(t)+dm(t), d3(t)=ds(t)+do(t), the filter transfer function is defined as H(f), the chip frequency is converted to D(f), and the spectrum of the output signal is The sampled spectrum is analyzed and the minimum mean square error is converted to E{|S(f)-H(f)·D(f)| 2}, let the partial derivative of this expression with respect to H(f) be zero, and obtain the transfer function of the optimal filter:

[0015] H(f)=S s (f) / (S s (f)+S x (f))

[0016] Where: S s (f) represents the power spectral density of the signal ds(t), S x (f) represents the power spectral density of noise dm(t), dn(t), and do(t), and H(f) is the ratio of the signal power spectrum to the total signal power spectrum.

[0017] The method for starting the A / D in S4 is as follows: when the adaptive flag is 0, first start to identify the noise by applying different intensities of Wiener filtering to the signal components of different frequency bands of the output digital quantity signal frequency with noise interference in the detection environment; in the filtered signal, based on the signal quality evaluation index, intelligently select the optimal sampling point to ensure the representativeness and accuracy of the sampling data; set the field switching power supply noise flag value to 1, the cable connector noise flag to 2, and the surrounding high-power equipment start-stop noise feature collection to 3; set the interference source with the greatest impact according to the identification, and apply the Wiener filter of this factor to the subsequent onboard sampling algorithm; when the adaptive Wiener filter selection flag is not 0, the A / D converted digital value is written into the register after subsequent secondary filtering for reading by the system program.

[0018] The method of adaptively selecting noise identification conditions in S5 is as follows: the standard digital signal quantity is restored by the corresponding Wiener filtering algorithm, the onboard processor sets the reading frequency of the A / D converted digital quantity, and limits the extreme value of the signal after Wiener filtering by arithmetic mean or median filtering, and writes the arithmetic mean or median into a register for reading by the back-end computer system software.

[0019] The S6 also includes: the signal processing program of the detection system reads the port value and performs corresponding conversion for the upper-layer application to call; the final digital signal is further smoothed and verified to ensure the continuity and consistency of the data and eliminate possible abnormal values.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] The present invention significantly improves the noise resistance and sampling stability after analog-to-digital conversion by introducing an improved Wiener filter, effectively solving the limitations of traditional methods in dealing with complex noise environments. The algorithm can not only maintain high signal fidelity, but also significantly improve the accuracy of data conversion and the overall robustness of the system, and is suitable for multiple fields such as high-precision measurement, signal processing, and communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0023] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0024] Figure 1 Schematic diagram of the detection method of the present invention.

[0025] Among them: 1 is the sensor analog signal, 2 is the interference analog signal, 3 is the mixed analog signal, 4 is A / D conversion, 5 is the adaptive Wiener filter selection mark, 6 is the digital characteristic noise recognition, 7 is the power supply noise Wiener filter, 8 is the cable connector noise filter, 9 is the high-power equipment interference noise Wiener filter, 10 is the secondary filtering, and 11 is the back-end signal quality feedback. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. These descriptions are only to further illustrate the features and advantages of the present invention, rather than to limit the claims of the present invention. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0027] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0028] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" 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 a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0029] The present invention proposes a sampling stability improvement algorithm based on improved Wiener filtering after the signal analog quantity is converted into digital quantity and adaptive noise interference conditions. The algorithm is implemented by the following steps:

[0030] Step 1: First, remove high-frequency noise and interference from the power supply system of the current sensor used for component detection and the A / D transmission line to the acquisition card to obtain the original analog signal and ensure that the effective information within the signal frequency band is not lost. The algorithm defines the target signal as s(t), the A / D conversion chip accuracy is a bit, the corresponding conversion voltage maximum value b, the fixed conversion ratio is k=2a / b, and the converted digital signal quantity is expressed as ds(t)=k*s(t);

[0031] Step 2: Collect and measure the noise environment on site, and collect the characteristics of the switch power supply noise, cable connector noise, and the start-stop noise of the surrounding high-power equipment. The algorithm defines the analog noise signal as m(t), n(t), o(t), and the digital signal as dm(t), dn(t), do(t);

[0032] Step 3: Perform Wiener filtering on the interference features to generate conversion digital signals d1(t)=ds(t)+dm(t), d2(t)=ds(t)+dm(t), d3(t)=ds(t)+do(t). Based on the traditional Wiener filtering, the filter transfer function is defined as H(f), and the chip frequency is converted to D(f). The spectrum of the output signal is The sampled spectrum is analyzed and the minimum mean square error is converted to E{|S(f)-H(f)·D(f)| 2}, let the partial derivative of this expression with respect to H(f) be zero, and we can get the transfer function of the optimal filter:

[0033] H(f)=S s (f) / (S s (f)+S x (f))

[0034] Where: S s (f) represents the power spectral density of the signal ds(t), S x (f) represents the power spectral density of noise dm(t), dn(t), and do(t), and H(f) is the ratio of the power spectrum of the signal to the total signal (signal plus noise).

[0035] Step 4. During the A / D startup process, set the adaptive Wiener filter selection flag to 0. When the adaptive flag is 0, first start to detect the output digital signal frequency with noise interference in the detection environment, and apply different intensities of Wiener filtering to the signal components of different frequency bands for noise identification. In the filtered signal, based on the signal quality evaluation indicators (such as signal-to-noise ratio, kurtosis, etc.), intelligently select the optimal sampling point to ensure the representativeness and accuracy of the sampling data. Set the on-site switching power supply noise flag value to 1, the cable connector noise flag to 2, and the surrounding high-power equipment start-stop noise feature collection to 3. According to the identification of the most influential setter interference source, and apply the factor Wiener filter to the subsequent onboard sampling algorithm. When the adaptive Wiener filter selection flag is not 0, the A / D converted digital value is written to the register for the system program to read after subsequent secondary filtering.

[0036] Step 5: According to the adaptive selection of noise recognition, the standard digital signal quantity is restored through the corresponding Wiener filtering algorithm. The onboard processor can set the A / D conversion digital quantity reading frequency, filter by arithmetic mean or median, limit the extreme value of the signal after Wiener filtering, and write the arithmetic mean or median into the register for reading by the back-end computer system software.

[0037] Step 6. Finally, a feedback mechanism is established to compare the sampled digital signal with the original analog signal, calculate the error, and feed the error information back to the improved Wiener filter for iterative optimization until the preset accuracy requirement is met. The signal processing program of the detection system reads the port value and performs corresponding conversion for the upper-level application to call. The final digital signal is further smoothed and verified to ensure the continuity and consistency of the data and eliminate possible outliers.

[0038] Example

[0039] This embodiment consists of sensor output analog signal, interference analog signal input, A / D analog-to-digital conversion, digital noise feature recognition, power supply noise Wiener filtering, cable connector noise filtering, high-power equipment interference noise Wiener filtering, secondary filtering, and back-end signal quality feedback optimization.

[0040] like Figure 1 As shown, the sensor analog signal 1 is an analog signal generated by the internal pressure transmitter chip of the detection sensor obtained in an interference-free environment. It is mixed with the interference analog signal 2 through the on-site communication cable of the detection equipment to form a mixed analog signal 3, and is input into the A / D conversion analog input terminal inside the high-speed acquisition control card.

[0041] The interference analog signal 2 is collected, analyzed and measured as the on-site interference source, and according to the Wiener filtering algorithm, it is improved and optimized to construct the power supply noise Wiener filter 7, cable connector noise filter 8, and high-power equipment interference noise Wiener filter 9. According to the back-end signal quality feedback 11 results, the power supply noise Wiener filter 7, cable connector noise filter 8, and high-power equipment interference noise Wiener filter 9 are continuously optimized.

[0042] The mixed analog signal 3 is input to the input end of the AD conversion module and converted into a digital output. Before the filtering is started, the adaptive Wiener filter selection identifier 5 is detected first. When the value of the adaptive Wiener filter selection identifier 5 is 0, the on-site digital characteristic noise identification 6 is started. After the main interference source is identified, the value of the adaptive Wiener filter selection identifier 5 is modified, and the algorithm initialization is completed. Then, the corresponding power supply noise Wiener filter 7, cable connector noise filter 8, and high-power equipment interference noise Wiener filter 9 are adaptively selected.

[0043] The digital signal filtered by the power supply noise Wiener filter 7, the cable connector noise filter 8 and the high-power equipment interference noise Wiener filter 9 enters the back-end secondary filter 10. The secondary filter 10 mainly adopts multi-frequency acquisition mean filtering or median filtering to achieve sampling stability, and passes through the back-end signal quality feedback 11.

[0044] The back-end signal quality feedback 11 evaluates the sampling stability, and continuously provides stable output if the stability meets the on-site detection needs. When the noise changes in the detection scene do not meet the needs of the detection system, the power supply noise Wiener filter 7, cable connector noise filter 8, high-power equipment interference noise Wiener filter 9 and secondary filter 10 are adjusted and optimized according to the optimization feedback requirements to achieve self-adaptation under various interference environment conditions.

[0045] Only the preferred embodiments of the present invention are described in detail above, but the present invention is not limited to the above embodiments. Various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention, and various changes should be included in the protection scope of the present invention.

Claims

1. A method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering, characterized in that: The following steps are involved: S1. Remove high-frequency noise and interference from the power supply system of the current sensor for component detection and the A / D transmission line to the acquisition card, obtain the original analog signal, ensure that the effective information within the signal frequency band is not lost, and convert it into a digital signal; S2. Collect and measure the noise environment on site, and collect the characteristics of the switch power supply noise, cable connector noise, and the start-stop noise of surrounding high-power equipment on site; S3, performing Wiener filtering and superposition on the interference characteristics to generate a converted digital signal; S4, during the A / D startup process, set the adaptive Wiener filter selection flag 0; S5, according to the adaptive selection noise recognition situation; S6. Establish a feedback mechanism to compare the sampled digital signal with the original analog signal, calculate the error, and feed the error information back to the improved Wiener filter for iterative optimization until the preset accuracy requirement is achieved.

2. The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering according to claim 1, characterized in that: The method for converting the digital signal quantity in S1 is as follows: the algorithm defines the target signal as s(t), the A / D conversion chip precision is a bit, the corresponding conversion voltage maximum value is b, the fixed conversion ratio is k=2a / b, and the converted digital signal quantity is expressed as ds(t)=k*s(t).

3. The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering according to claim 1, characterized in that: The analog noise signal of the on-site noise environment in S2 is defined as m(t), n(t), o(t), and the digital noise signal is defined as dm(t), dn(t), do(t).

4. The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering according to claim 1, characterized in that: The converted digital signals in S3 are d1(t)=ds(t)+dm(t), d2(t)=ds(t)+dm(t), d3(t)=ds(t)+do(t), the filter transfer function is defined as H(f), the chip frequency is converted to D(f), and the spectrum of the output signal is The sampled spectrum is analyzed and the minimum mean square error is converted to E{|S(f)-H(f)·D(f)| 2 }, let the partial derivative of this expression with respect to H(f) be zero, and obtain the transfer function of the optimal filter: H(f)=S s (f) / (S s (f)+S x (f)) Where: S s (f) represents the power spectral density of the signal ds(t), S x (f) represents the power spectral density of noise dm(t), dn(t), and do(t), and H(f) is the ratio of the signal power spectrum to the total signal power spectrum.

5. The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering according to claim 1, characterized in that: The method for starting the A / D in S4 is as follows: when the adaptive flag is 0, the output digital signal frequency with noise interference in the detection environment is firstly applied to the signal components of different frequency bands by Wiener filtering with different strengths for noise identification, and in the filtered signal, based on the signal quality evaluation index, the optimal sampling point is intelligently selected to ensure the representativeness and accuracy of the sampling data; Set the on-site switching power supply noise flag value to 1, the cable connector noise flag to 2, and the surrounding high-power equipment start-stop noise feature collection to 3; According to the identification of the interference source with the greatest impact, the Wiener filter of this factor is applied to the subsequent onboard sampling algorithm; when the adaptive Wiener filter selection flag is not 0, the A / D converted digital value is written into the register after subsequent secondary filtering for the system program to read.

6. The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering according to claim 1, characterized in that: The method of adaptively selecting noise identification conditions in S5 is as follows: the standard digital signal quantity is restored by the corresponding Wiener filtering algorithm, the onboard processor sets the reading frequency of the A / D converted digital quantity, and limits the extreme value of the signal after Wiener filtering by arithmetic mean or median filtering, and writes the arithmetic mean or median into a register for reading by the back-end computer system software.

7. The method for improving the sampling stability of high-speed analog-to-digital conversion based on improved Wiener filtering according to claim 1, characterized in that: The S6 also includes: the signal processing program of the detection system reads the port value and performs corresponding conversion for the upper-layer application to call; the final digital signal is further smoothed and verified to ensure the continuity and consistency of the data and eliminate possible abnormal values.

Images