Signal adaptive detection method, device, equipment, storage medium and program product

Abstract

The application discloses a signal self-adaptive detection method, which comprises the following steps: obtaining a to-be-processed signal, segmenting the to-be-processed signal based on a preset unit signal width to obtain a plurality of to-be-processed segmented signals, setting a time window, performing delay autocorrelation processing on each to-be-processed segmented signal in the time window to obtain a delay autocorrelation calculation value, performing histogram statistics on the delay autocorrelation calculation values of all the to-be-processed segmented signals in the time window to determine whether the signal type in the time window is unique, performing frequency domain analysis on the to-be-processed segmented signals in the time window to determine the signal type in the time window, determining whether there is a pulse start position and / or a pulse end position of the to-be-processed signal, and counting the pulse position detection results in each step sliding time window to obtain a signal detection result. The application improves the applicability of signal detection.

Description

Technical Field

[0001] This application relates to the field of signal detection technology, and in particular to signal adaptive detection methods, devices, equipment, storage media, and program products. Background Technology

[0002] Signal detection is the foundation of signal processing; it involves detecting the presence of a useful signal within noise. All signal processing is based on the premise that a useful signal exists. Currently, common signal detection methods mainly include those based on time-domain statistics and frequency-domain statistics. Time-domain statistics commonly involve signal energy statistics, while frequency-domain statistics include power spectrum statistics and amplitude spectrum statistics. Time-domain energy statistics detection methods calculate a detection threshold based on the statistical distribution of signal energy over a period of time, determining if the noise energy amplitude is lower than the useful signal's energy amplitude during that period. Similarly, frequency-domain statistical detection, based on power spectrum and amplitude spectrum, calculates a threshold based on the differences in power and amplitude between noise and the useful signal over a period of time, thus determining if a useful signal is detected.

[0003] The aforementioned common signal detection methods statistically analyze the signal distribution over a cumulative time period, assuming the presence of both noise and useful signals within that period. They statistically analyze the signal energy distribution within the current time period, determine a detection threshold, and use this threshold to calculate the useful signal within that time period. This approach is not adaptable to changes in different signal types and may result in inaccurate detection of the signal start / end position. Summary of the Invention

[0004] The main objective of this application is to provide a signal adaptive detection method, apparatus, device, storage medium, and program product, which aims to solve the technical problem of inaccurate detection of the start / end position of signals in related technologies.

[0005] To achieve the above objectives, this application proposes a signal adaptive detection method, the method comprising: Acquire the signal to be processed, and segment the signal to be processed based on a preset unit signal width to obtain multiple segments of the signal to be processed. Set a time window, and for each segment of the signal to be processed within the time window, perform delayed autocorrelation processing on the segment of the signal to be processed to obtain the delayed autocorrelation value. Histogram statistics are performed on the calculated values ​​of the delayed autocorrelation of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique. If the signal type is unique within the time window, then frequency domain analysis is performed on the segmented signal to be processed within the time window to determine the signal type within the time window. Based on the signal detection threshold corresponding to the signal type distribution within the time window, pulse detection is performed on the sliding time window to determine whether there is a pulse start position and / or pulse end position of the signal to be processed; By using a sliding time window, the pulse position detection results within each sliding time window are statistically analyzed to determine the start and / or end positions of all pulses in the signal to be processed, thereby obtaining the signal detection results.

[0006] In one embodiment, the step of performing histogram statistics on the delayed autocorrelation values ​​of all segments of signals to be processed within a time window to determine whether the signal type within the time window is unique includes: Determine the maximum and minimum values ​​of the delayed autocorrelation calculations for all segmented signals to be processed within the time window; Using the minimum value as the starting point of the statistical interval and the maximum value as the ending point of the statistical interval, the interval to be evaluated is divided into M non-overlapping statistical regions at equal intervals. The time window is traversed to calculate the delayed autocorrelation values ​​of all segments of the signal to be processed. The interval distribution of the delayed autocorrelation values ​​of the segments of the signal to be processed is statistically analyzed. The number of segments of the signal to be processed falling into each statistical region is determined, and a histogram is constructed. The data intervals of the histogram correspond to the statistical regions, and the data frequency in each data interval corresponds to the number of segments of the signal to be processed falling into the corresponding statistical region. The number of peaks in the histogram and the numerical difference between the peaks of adjacent peaks are counted. If the number of peaks is 2 and the numerical difference is not greater than the first judgment threshold, then the signal type within the time window is determined to be unique.

[0007] In one embodiment, if the signal type is unique within a time window, the step of performing frequency domain analysis on the segmented signal to be processed within the time window to determine the signal type within the time window includes: Perform Fast Fourier Transform on all segmented signals to be processed within the time window to obtain the frequency domain data sequence to be processed; The distribution characteristics of the frequency domain data to be processed are statistically analyzed, and the signal type within the time window is determined based on the statistical results; the signal type includes useful signals and noise signals.

[0008] In one embodiment, the step of performing distribution characteristic statistics on the frequency domain data to be processed and determining the signal type within a time window based on the statistical results includes: The frequency domain data sequence is divided into three non-overlapping frequency domain intervals; For any two frequency domain intervals, calculate the difference between the maximum value in the first frequency domain interval and the maximum value in the second frequency domain interval to obtain the first difference, the second difference, and the third difference; The maximum difference is determined from the first, second, and third differences. If the maximum difference is greater than the second judgment threshold, the signal type within the time window is determined to be a useful signal.

[0009] In one embodiment, the step of performing pulse detection on the sliding time window based on the signal detection threshold corresponding to the signal type distribution within the time window, and determining whether there is a pulse start position and / or pulse end position of the signal to be processed, includes: Determine the signal detection threshold corresponding to the signal type distribution within the time window; The delayed autocorrelation values ​​of each segment of signal to be processed within the time window are compared sequentially with the signal detection threshold in chronological order. If the calculated value of the delayed autocorrelation of the segmented signal to be processed is greater than the signal detection threshold, then the first cumulative value is incremented by 1 and the second cumulative value is set to zero. If the calculated value of the delayed autocorrelation of the segmented signal to be processed is not greater than the signal detection threshold, then the second cumulative value is incremented by 1 and the first cumulative value is set to zero. If the first accumulated value reaches the starting position judgment value and the pulse flag state is the first state, then based on the position of the segment signal to be processed corresponding to the time when the starting position judgment value is reached, the pulse start position is determined and the pulse flag is set to the second state. If the second accumulated value reaches the end position judgment value and the pulse flag state is the second state, then the pulse end position is determined based on the position of the segment signal to be processed when the end position judgment value is reached.

[0010] In one embodiment, the steps of sliding a time window, statistically analyzing the pulse position detection results within each sliding time window, determining the start and / or end positions of all pulses in the signal to be processed, and obtaining the signal detection results include: The sliding time window is used to determine whether there is a pulse start position and / or pulse end position in the next time window, until the signal to be processed is traversed. For each time window, the initial state of the pulse flag corresponding to the current time window inherits the final state of the pulse flag of the previous time window.

[0011] Secondly, to achieve the above objectives, this application further provides a signal adaptive detection device, the device comprising: The acquisition module is used to acquire the signal to be processed, and to segment the signal to be processed based on a preset unit signal width to obtain multiple segments of the signal to be processed. The autocorrelation value calculation module is used to set a time window, and for each segment of the signal to be processed within the time window, it performs delayed autocorrelation processing on the segment of the signal to be processed to obtain the delayed autocorrelation value. The first signal type determination module is used to perform histogram statistics on the delay autocorrelation calculation values ​​of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique. The second signal type determination module is used to perform frequency domain analysis on the segmented signal to be processed within the time window if the signal type is unique within the time window, and to determine the signal type within the time window. The first position detection module is used to perform pulse detection on the sliding time window by combining the signal detection threshold corresponding to the signal type distribution within the time window, and to determine whether there is a pulse start position and / or pulse end position of the signal to be processed. The second position detection module is used to slide the time window, count the pulse position detection results within each step of the sliding time window, determine the start position and / or end position of all pulses in the signal to be processed, and obtain the signal detection result.

[0012] Thirdly, to achieve the above objectives, this application further provides a signal adaptive detection device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the above-described signal adaptive detection method.

[0013] Fourthly, to achieve the above objectives, this application further provides a storage medium, which is a computer-readable storage medium, and stores a computer program thereon. When the computer program is executed by a processor, it implements the steps of the above-described adaptive signal detection method.

[0014] Fifthly, to achieve the above objectives, this application further provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the above-described adaptive signal detection method.

[0015] One or more technical solutions proposed in this application have at least the following technical effects: This application achieves adaptive signal type identification by segmenting the signal to be processed into segments with a preset unit signal width and combining sliding time windows, delay autocorrelation analysis, time-domain histogram statistics, and frequency-domain feature discrimination. It dynamically determines the detection threshold based on different signal types, thus avoiding the problem of insufficient adaptability to signal variations in traditional fixed threshold or single time-domain statistical methods. Simultaneously, by performing continuous threshold decision-making and pulse state tracking on the calculated delay autocorrelation value, the start and end positions of the pulse signal can be determined more accurately. Combined with continuous detection of the entire signal to be processed using a sliding time window, real-time updating of the detection threshold and stable detection under complex noise environments are achieved, effectively improving the accuracy, adaptability, and real-time performance of pulse detection in different signal scenarios. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the adaptive signal detection method in the embodiments of this application.

[0019] Figure 2 This is a schematic diagram of the module connections for a signal adaptive detection device.

[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0023] Specifically, embodiments of this application provide a signal adaptive detection method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the adaptive signal detection method of this application.

[0024] In this embodiment, the signal adaptive detection method includes steps S10 to S60: Step S10: Obtain the signal to be processed, and segment the signal to be processed based on a preset unit signal width to obtain multiple segments of the signal to be processed. Step S20: Set a time window, and perform delayed autocorrelation processing on each segment of the signal to be processed within the time window to obtain the calculated value of delayed autocorrelation. Step S30: Perform histogram statistics on the calculated delay autocorrelation values ​​of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique. Step S40: If the signal type is unique within the time window, then perform frequency domain analysis on the segmented signal to be processed within the time window to determine the signal type within the time window. Step S50: Combine the signal detection threshold corresponding to the signal type distribution within the time window, perform pulse detection on the sliding time window to determine whether there is a pulse start position and / or pulse end position of the signal to be processed; Step S60: Slide the time window, statistically analyze the pulse position detection results within each sliding time window, determine the start position and / or end position of all pulses in the signal to be processed, and obtain the signal detection results.

[0025] In one feasible implementation, step S30 includes steps A10 to A40: Step A10: Determine the maximum and minimum values ​​of the calculated delay autocorrelation values ​​of all segmented signals to be processed within the time window.

[0026] Step A20: Using the minimum value as the starting point of the statistical interval and the maximum value as the ending point of the statistical interval, the interval to be evaluated is divided into M non-overlapping statistical regions at equal intervals.

[0027] Step A30: Iterate through the delayed autocorrelation values ​​of all segmented signals to be processed within the time window, statistically analyze the interval distribution of the delayed autocorrelation values ​​of the segmented signals to be processed, determine the number of segmented signals to be processed falling into each statistical region, and construct a histogram. The data intervals of the histogram correspond to the statistical regions, and the data frequency in each data interval corresponds to the number of segmented signals to be processed falling into the corresponding statistical region.

[0028] Step A40: Count the number of peaks in the histogram and the numerical difference between the peaks of adjacent peaks. If the number of peaks is 2 and the numerical difference is not greater than the first judgment threshold, then the signal type within the time window is determined to be unique.

[0029] Step S40 includes steps B10 to B20: Step B10: Perform Fast Fourier Transform on all segments of the signal to be processed within the time window to obtain the frequency domain data sequence to be processed.

[0030] Step B20: Perform distribution characteristic statistics on the frequency domain data to be processed, and determine the signal type within the time window based on the statistical results; the signal type includes useful signals and noise signals.

[0031] Step B20 includes steps B21 to B23: Step B21: Divide the frequency domain data sequence into three non-overlapping frequency domain intervals; Step B22: For any two frequency domain intervals, calculate the difference between the maximum value in the first frequency domain interval and the maximum value in the second frequency domain interval, and obtain the first difference, the second difference, and the third difference. Step B23: Determine the maximum difference from the first difference, the second difference, and the third difference. If the maximum difference is greater than the second judgment threshold, then determine that the signal type within the time window is a useful signal.

[0032] Step S50 includes steps C10 to C60: Step C10: Determine the signal detection threshold corresponding to the signal type distribution within the time window.

[0033] Step C20: Compare the calculated delay autocorrelation values ​​of each segment of signal to be processed within the time window with the signal detection threshold in chronological order.

[0034] Step C30: If the calculated value of the delayed autocorrelation of the segmented signal to be processed is greater than the signal detection threshold, then the first cumulative value is incremented by 1 and the second cumulative value is set to zero.

[0035] Step C40: If the calculated value of the delayed autocorrelation of the segmented signal to be processed is not greater than the signal detection threshold, then increment the second cumulative value by 1 and set the first cumulative value to zero.

[0036] Step C50: If the first accumulated value reaches the starting position judgment value and the pulse flag state is the first state, then based on the position of the segment signal to be processed corresponding to the time the starting position judgment value is reached, the pulse start position is determined and the pulse flag is set to the second state.

[0037] Step C60: If the second accumulated value reaches the end position judgment value and the pulse flag state is the second state, then the pulse end position is determined based on the position of the segment signal to be processed when the end position judgment value is reached.

[0038] Step S60 includes step D10: Step D10: Slide the time window to determine whether there is a pulse start position and / or pulse end position in the next time window, until the signal to be processed is traversed; wherein, for each time window, the initial state of the pulse flag corresponding to the current time window inherits the final state of the pulse flag of the previous time window.

[0039] Specifically, the signal to be processed refers to the raw time-domain signal to be detected, which may contain both useful and noise signals. The preset unit signal width is the smallest processing unit pre-defined based on the typical pulse width, sampling rate, or detection accuracy of the pulse signal to be detected. It is used to divide the continuous signal to be processed into multiple segments of uniform length, ensuring that each segment corresponds to a fixed time range of sampled data. By standardizing the segmented processing of the signal to be processed, the data scale of subsequent calculations can be reduced, improving real-time processing efficiency. Furthermore, it establishes a unified data processing granularity for delay autocorrelation calculation, time window statistics, and pulse position determination, ensuring that subsequent processing steps are performed with a consistent unit signal length, thus improving the stability and consistency of the entire detection process.

[0040] The time window is a sliding analysis region covering multiple segments of the signal to be processed. Its length is typically greater than the length of a single segment and it moves continuously according to a preset sliding step size. Delayed autocorrelation processing involves performing correlation operations between each segment of the signal to be processed and its delayed version to obtain the corresponding delayed autocorrelation value. Since useful signals usually have a certain time correlation, while noise signals are highly random, after delayed autocorrelation processing, the calculated delayed autocorrelation value for the useful signal is usually significantly higher than that for the noise signal. This further amplifies the numerical difference between the two, improves the separability of signal and noise under low signal-to-noise ratio conditions, and provides more obvious data characteristics for subsequent statistical analysis.

[0041] The calculated delayed autocorrelation values ​​of all segmented signals to be processed within the current time window are used as statistical objects. A histogram is constructed to statistically analyze the distribution of these calculated values ​​across different numerical intervals. Each statistical region in the histogram corresponds to an interval of calculated delayed autocorrelation values, and its frequency represents the number of segmented signals to be processed falling within that interval. Based on the number of peaks in the histogram and the distribution relationship between different peaks, it can be determined whether useful and noise signals coexist within the current time window. When two distinct peaks are formed in the histogram, it indicates that two types of signals coexist within the window; when only a single peak is formed, it indicates that the window contains only one type of signal, meaning the signal type within the current time window is unique. Using delayed autocorrelation statistics combined with histogram distribution analysis, rather than directly relying on energy statistics, can more accurately reflect signal distribution characteristics and improve the reliability of signal type identification in complex noisy environments.

[0042] When it is determined that the signal type within the current time window is unique, time-domain statistics alone cannot further distinguish whether the entire signal within the window is useful or noise. Therefore, a Fast Fourier Transform (FFT) is performed on the segmented signal to be processed within the time window to obtain the corresponding frequency-domain data. The distribution characteristics of this frequency-domain data are then used to complete a secondary discrimination. Since continuous useful signals typically have a relatively obvious concentrated spectral energy region, while noise signals have a relatively uniform spectral distribution, frequency-domain statistics can effectively distinguish between the two types, both exhibiting a single-peak distribution. This avoids missed detections of continuous useful signals due to similarity between time-domain statistical results and noise, thus improving the accuracy of signal type discrimination in different signal scenarios.

[0043] Based on the signal type determined within the current time window, the corresponding signal detection threshold is selected. The calculated delay autocorrelation value of each segment of the signal to be processed is then compared with the signal detection threshold. The status of segments continuously exceeding or falling below the threshold is statistically analyzed using the first cumulative value, the second cumulative value, and a pulse flag. When the threshold is continuously exceeded to reach the start position decision value, the pulse start position is determined; when the threshold is continuously fallen to reach the end position decision value, the pulse end position is determined. Because continuous cumulative decision-making is used instead of single-point comparison, misjudgments caused by noise fluctuations can be effectively suppressed, making the detection of pulse start and end positions more stable and accurate. Furthermore, the adaptive signal detection thresholds corresponding to different signal types improve the adaptability of the detection threshold to different signal environments.

[0044] After completing the analysis of the current time window, the time window is slid along according to a preset sliding step size, and the delay autocorrelation statistics, signal type identification, and pulse position detection in subsequent time windows are performed. Simultaneously, the pulse flag state at the end of the previous time window is inherited to the next time window, ensuring that pulse signals existing continuously across windows maintain a continuous detection state, thus avoiding pulse signal truncation or repeated detection due to window switching. By continuously sliding the time window to traverse the entire signal to be processed, and combining the pulse start and end positions obtained in each sliding time window, a complete signal detection result is finally obtained.

[0045] Understandably, this embodiment uses a sliding window approach. After performing sliding window autocorrelation processing on the signal in the time domain, it performs histogram statistics on the processed time-domain signal and makes an initial judgment based on the distribution pattern of the histogram to determine whether there is a useful signal in the current time period. If a useful signal exists, a detection threshold is calculated based on the histogram statistics of the useful signal and noise. Otherwise, it enters the second judgment stage. By statistically analyzing the frequency domain characteristics of the signal in the current window length, it is determined whether it is entirely a useful signal or noise. If it is entirely a useful signal, the window length data is directly detected as a signal segment, and a detection threshold is calculated based on the useful signal segment. Otherwise, it is a noise signal, and a detection threshold is calculated based on the statistical characteristics of the noise signal segment. Based on the decision threshold calculated by the current sliding window, the signal in the next sliding window segment is detected, and the next stage of sliding window calculation and signal detection is performed.

[0046] By employing autocorrelation calculations, this method addresses the problem in current detection technologies where the energy difference between the useful signal and noise is small under low signal-to-noise ratio (SNR) conditions, making it impossible to calculate a suitable detection threshold. Based on the irregular nature of noise and the temporal regularity of the useful signal, a time-delayed autocorrelation method increases the amplitude of the useful signal after autocorrelation, while the amplitude of the noise remains relatively unchanged. This further amplifies the amplitude difference between the useful signal and noise, thereby resolving the inaccuracy issue in low SNR detection.

[0047] By employing a sliding window and histogram statistical method based on autocorrelation values, real-time detection can be guaranteed, and the detection threshold can be dynamically updated. This addresses the problem in real-world scenarios where the energy amplitudes of noise and useful signals can vary significantly, making a fixed detection threshold unsuitable and enabling real-time adaptive threshold updates. Within each sliding window period, the detection threshold is updated in real-time based on the histogram statistical results. The detection threshold of the current window is used to detect the signal in the next sliding window, achieving signal detection with adaptively updated thresholds for different scenarios.

[0048] By employing a joint time-frequency domain statistical method, this approach addresses the issue of missed detections of continuous useful signals when time-domain histogram statistics are insufficient to determine the signal distribution characteristics within the statistical timeframe. Based on the frequency domain distribution characteristics of useful signals and noise, the time-domain histogram statistical results can be supplemented with additional detection. This leverages frequency domain distribution characteristics to address situations where only useful signals or noise appear within the entire segment, thus solving the challenge of detecting continuous signals within the statistical timeframe and resulting in more accurate signal detection.

[0049] For example, this example provides an adaptive signal detection method based on joint statistical autocorrelation in the time and frequency domains. To ensure real-time signal detection, this example employs a sliding window approach, where the step size of the sliding window is smaller than the window length. When a useful signal appears, the sliding window always contains both noise and the useful signal, ensuring the effectiveness and accuracy of subsequent histogram statistics and signal detection. Furthermore, the sliding window approach ensures effective signal detection within a very short time after the useful signal appears. Considering the difference in energy amplitude between noise and signal, a histogram statistical method is used. According to the principle of histogram statistics, noise and the useful signal will only produce two peaks in the histogram, and the presence of a useful signal is determined based on the interval between the two peaks.

[0050] An adaptive signal detection method combining sliding window and histogram statistics was designed based on the distribution characteristics of signal and noise, realizing adaptive detection throughout the entire signal detection process. It eliminates the need to manually set the signal detection threshold and minimizes the impact of environmental noise on the overall energy amplitude of the signal, enabling adaptive signal detection for different signal environments. To ensure the accuracy of signal detection, this example employs a two-stage decision method based on joint time-frequency domain statistics. In the first decision, after applying an adaptive signal detection method using a sliding window and time-domain histogram statistics, the signal within the sliding window length is classified into two cases: one where useful signals and noise coexist, and another. In the second decision, the other cases after the first decision are further categorized into completely continuous useful signals and completely noisy signals based on their statistical characteristics in the frequency domain. This two-stage decision method based on joint time-frequency domain statistics categorizes the signal distribution in real-world scenarios into three types, employing different adaptive threshold calculation methods for each type. This ensures the real-time performance and accuracy of signal detection and addresses the issue of missed detections when continuous useful signals appear within a short period, which occurs when threshold calculation is performed solely based on histogram statistics after signal autocorrelation.

[0051] Existing signal detection methods require calculating the signal detection threshold based on the statistical characteristics of a known noise segment; or calculating the threshold after accumulating noise over a period of time, when both noise and useful signals exist within that period, thus obtaining the signal detection threshold for the current time period. In practical scenarios, these methods cannot perform threshold calculations with known noise segments, nor can they detect signals within continuous useful signal segments. Furthermore, they cannot perform signal detection after accumulating data over a period of time, failing to meet the real-time signal detection requirements of real-world scenarios.

[0052] This example, based on the energy distribution characteristics of noise and useful signals in the time-frequency domain, employs a joint time-frequency domain autocorrelation method. This further amplifies the difference between the useful signal and noise components in the time domain. By using a delayed autocorrelation method, signal detection can be performed even under low signal-to-noise ratio conditions, making this invention applicable to more scenarios. The detection threshold adaptive update method based on autocorrelation value histogram statistics employed in this invention, after determining the presence of a useful signal through histogram statistics, calculates a threshold for the signal within the current window. This ensures that after each detected pulse signal, the threshold automatically calculates the noise value after the pulse ends to determine the detection threshold for the next pulse. This allows the detection threshold to be adaptively updated in real time according to the current signal environment, effectively reducing the impact of changes in the signal environment on signal detection.

[0053] This example further distinguishes useful signals from noise in the frequency domain by leveraging frequency domain distribution characteristics. Based on the different frequency domain distribution characteristics of useful signals and noise, different threshold calculation methods are adopted in the time domain for different situations, solving the problem of missed detection of continuous useful signals and improving the accuracy of signal detection.

[0054] Specifically, the first step is to set the sliding window length. The sliding window length is set based on the pulse width of the pulse signal to be detected. , Given the received raw signal, the sliding window slides according to the sliding window step n each time.

[0055] Within the set window length, perform delayed autocorrelation processing to obtain the autocorrelation-processed value. .

[0056] For the received raw signal Its complex expression is: in, Indicates taking the real part, To obtain the imaginary part, It is the imaginary unit.

[0057] Time-delayed autocorrelation is the process of converting the original signal into a time-delayed autocorrelation signal. Multiplying by itself with a delay, the expression is as follows: It can be calculated using piecewise sliding, and its expression is as follows: Perform histogram statistics on the data within the current window after time-delayed autocorrelation processing.

[0058] Divide the valid interval of the data within the current window after delayed autocorrelation processing into equal parts. Obtain the interval between the minimum and maximum values ​​of the data: Divide this interval into M equal parts, with each part having a value of R. Then: Find the set of the i-th interval. According to step 3.1, each interval, from smallest to largest, can be represented as follows: Therefore, it can be defined The set of the i-th interval of the histogram: in, For the current window The value after time-delayed autocorrelation processing.

[0059] Count the number of autocorrelation values ​​in each interval. Iterate through the data in the current window after delayed autocorrelation processing. The initial count for each interval is 0. If an autocorrelation value falls in the i-th interval, the count in that interval is incremented by 1, until the entire set of autocorrelation values ​​has been traversed.

[0060] Count the number of peaks in the histogram and the numerical difference between the peaks. After the histogram statistics in steps 3.1-3.3, count the number of peaks in the histogram and the numerical difference between the peaks of any two adjacent peaks: in, The value represented by the peak of the i-th wave after time-delay autocorrelation processing.

[0061] If the histogram has only two peaks, and the numerical difference between the peaks of these two peaks is... The threshold value for determining the presence or absence of a signal is greater than the set threshold value. If the signal is positive, then there is a useful signal and noise in the current window, and the subsequent signal detection threshold is calculated; otherwise, there is no useful signal, and the next stage of frequency domain decision is performed.

[0062] When the current sliding window segment detects the absence of both useful signal and noise, a Fast Fourier Transform is performed on the data within the current sliding window segment. Assume the data length in the time domain is N: The frequency domain data after Fast Fourier Transform are statistically analyzed to determine whether the signal within the current window is a continuous useful signal or noise. Divide the frequency domain data into three intervals according to the length N: [0, N / 3], [N / 3, 2N / 3], and [2N / 3, N]. Calculate the maximum value for each interval. Calculate the maximum difference between the maximum values ​​of the three intervals: When the maximum difference Exceeding the threshold If the signal is positive, then there is a continuous useful signal within that sliding window length; otherwise, it is all noise.

[0063] Calculate the signal detection threshold When a useful signal is detected within the sliding window, the signal detection threshold is calculated using the following formula: After testing, if the current sliding window length contains only continuous useful signals, then the threshold calculation formula is: After testing, the entire length of the current sliding window is filled with noise. Therefore, the threshold calculation formula is: The delay autocorrelation value is compared with the signal detection threshold. When the delay autocorrelation value is greater than the signal detection threshold: For the cumulative number of signals exceeding the detection threshold Perform the accumulation operation: when When the number of signals detected is less than the cumulative value of the signal detection threshold, it will be less than the cumulative value of the signal detection threshold Reset to 0: The delayed autocorrelation value is compared with the signal detection threshold. When the delayed autocorrelation value is less than or equal to the signal detection threshold: For the cumulative number of signals less than the signal detection threshold Perform the accumulation operation: when When the number of signals exceeds the cumulative value of the signal detection threshold, it will be greater than the cumulative value of the threshold. Reset to 0: When the cumulative number of signals detected exceeds the threshold value Greater than the set pulse signal start judgment value When the pulse flag is 0, it can be determined that the pulse signal has started. The current time is the start time of the pulse signal, and the pulse flag is set to 1 so that the flag can be changed for subsequent determination of the pulse signal's end time.

[0064] When the number of signals detected is less than the cumulative value of the signal detection threshold Greater than the set pulse signal start judgment value When the pulse flag is 1, it can be determined that the pulse signal has ended, the current time is the end time of the pulse signal, and the pulse flag is set to 0, so that the flag can be changed to determine the start time of the next pulse signal.

[0065] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the signal adaptive detection method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0066] This application also provides a signal adaptive detection device, please refer to... Figure 2 The signal adaptive detection device includes: The acquisition module 10 is used to acquire the signal to be processed, and to segment the signal to be processed based on a preset unit signal width to obtain multiple segments of the signal to be processed. The autocorrelation value calculation module 20 is used to set a time window, and for each segment of the signal to be processed within the time window, perform delayed autocorrelation processing on the segment of the signal to be processed to obtain the delayed autocorrelation value. The first signal type determination module 30 is used to perform histogram statistics on the delay autocorrelation calculation values ​​of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique. The second signal type determination module 40 is used to perform frequency domain analysis on the segmented signal to be processed within the time window if the signal type is unique within the time window, and to determine the signal type within the time window. The first position detection module 50 is used to perform pulse detection on the sliding time window in combination with the signal detection threshold corresponding to the signal type distribution within the time window, and to determine whether there is a pulse start position and / or pulse end position of the signal to be processed. The second position detection module 60 is used to slide the time window, count the pulse position detection results within each step of the sliding time window, determine the start position and / or end position of all pulses in the signal to be processed, and obtain the signal detection result.

[0067] The signal adaptive detection device provided in this application, employing the signal adaptive detection method in the above embodiments, can solve the technical problem of low pulse detection efficiency in related technologies. Compared with related technologies, the beneficial effects of the signal adaptive detection device provided in this application are the same as those of the signal adaptive detection method provided in the above embodiments, and other technical features in the signal adaptive detection device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0068] This application provides a signal adaptive detection device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the signal adaptive detection method described in the above embodiments.

[0069] The signal adaptive detection device provided in this application, employing the signal adaptive detection method described in the above embodiments, can solve the technical problem of low pulse detection efficiency in related technologies. Compared with related technologies, the beneficial effects of the signal adaptive detection device provided in this application are the same as those of the signal adaptive detection method described in the above embodiments, and other technical features of this signal adaptive detection device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0070] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0071] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0072] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the signal adaptive detection method in the above embodiments.

[0073] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described adaptive signal detection method, thereby solving the technical problem of low pulse detection efficiency in related technologies. Compared with related technologies, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the adaptive signal detection method provided in the above embodiments, and will not be repeated here.

[0074] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the signal adaptive detection method described above.

[0075] The computer program product provided in this application can solve the technical problem of low pulse detection efficiency in related technologies. Compared with related technologies, the beneficial effects of the computer program product provided in this application are the same as those of the signal adaptive detection method provided in the above embodiments, and will not be repeated here.

[0076] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A signal adaptive detection method, characterized in that, The method includes: Acquire the signal to be processed, and segment the signal to be processed based on a preset unit signal width to obtain multiple segments of the signal to be processed. Set a time window, and perform delayed autocorrelation processing on each segment of the signal to be processed within the time window to obtain the delayed autocorrelation calculation value. Histogram statistics are performed on the calculated delay autocorrelation values ​​of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique. If the signal type is unique within the time window, then frequency domain analysis is performed on the segmented signal to be processed within the time window to determine the signal type within the time window; Based on the signal detection threshold corresponding to the signal type distribution within the time window, pulse detection is performed on the sliding time window to determine whether there is a pulse start position and / or pulse end position of the signal to be processed; By using a sliding time window, the pulse position detection results within each sliding time window are statistically analyzed to determine the start and / or end positions of all pulses in the signal to be processed, thereby obtaining the signal detection results.

2. The signal adaptive detection method as described in claim 1, characterized in that, The step of performing histogram statistics on the calculated delay autocorrelation values ​​of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique includes: Determine the maximum and minimum values ​​of the delayed autocorrelation calculations for all segmented signals to be processed within the time window; Using the minimum value as the starting point of the statistical interval and the maximum value as the ending point of the statistical interval, the interval to be evaluated is divided into M non-overlapping statistical regions at equal intervals. The time window is traversed to calculate the delayed autocorrelation values ​​of all segments of the signal to be processed, and the interval distribution of the calculated delayed autocorrelation values ​​of the segments of the signal to be processed is statistically analyzed. The number of segments of the signal to be processed falling into each statistical region is determined, and a histogram is constructed. The data intervals of the histogram correspond to the statistical regions, and the data frequency in each data interval corresponds to the number of segments of the signal to be processed falling into the corresponding statistical region. The number of peaks in the histogram and the numerical difference between the peaks of adjacent peaks are counted. If the number of peaks is 2 and the numerical difference is not greater than the first judgment threshold, then the signal type within the time window is determined to be unique.

3. The signal adaptive detection method as described in claim 1, characterized in that, If the signal type is unique within the time window, the step of performing frequency domain analysis on the segmented signal to be processed within the time window to determine the signal type within the time window includes: Perform Fast Fourier Transform on all segmented signals to be processed within the time window to obtain the frequency domain data sequence to be processed; The distribution characteristics of the frequency domain data to be processed are statistically analyzed, and the signal type within the time window is determined based on the statistical results; the signal type includes useful signals and noise signals.

4. The signal adaptive detection method as described in claim 3, characterized in that, The step of performing distribution characteristic statistics on the frequency domain data to be processed and determining the signal type within the time window based on the statistical results includes: The frequency domain data sequence is divided into three non-overlapping frequency domain intervals; For any two frequency domain intervals, calculate the difference between the maximum value in the first frequency domain interval and the maximum value in the second frequency domain interval to obtain the first difference, the second difference, and the third difference; The maximum difference is determined from the first difference, the second difference, and the third difference. If the maximum difference is greater than the second judgment threshold, the signal type within the time window is determined to be a useful signal.

5. The signal adaptive detection method as described in claim 1, characterized in that, The step of performing pulse detection on the sliding time window based on the signal detection threshold corresponding to the signal type distribution within the time window, and determining whether there is a pulse start position and / or pulse end position of the signal to be processed, includes: Determine the signal detection threshold corresponding to the signal type distribution within the time window; The delayed autocorrelation values ​​of each segment of signal to be processed within the time window are compared sequentially with the signal detection threshold in chronological order. If the calculated value of the delayed autocorrelation of the segmented signal to be processed is greater than the signal detection threshold, then the first cumulative value is incremented by 1 and the second cumulative value is set to zero. If the calculated value of the delayed autocorrelation of the segmented signal to be processed is not greater than the signal detection threshold, then the second cumulative value is incremented by 1 and the first cumulative value is set to zero. If the first accumulated value reaches the starting position judgment value and the pulse flag state is the first state, then based on the position of the segment signal to be processed corresponding to the time when the starting position judgment value is reached, the pulse starting position is determined and the pulse flag is set to the second state. If the second accumulated value reaches the end position judgment value and the pulse flag state is the second state, then the pulse end position is determined based on the position of the segment signal to be processed when the end position judgment value is reached.

6. The signal adaptive detection method as described in claim 5, characterized in that, The sliding time window, which statistically analyzes the pulse position detection results within each step of the sliding time window, determines the start and / or end positions of all pulses in the signal to be processed, and the steps to obtain the signal detection results include: A sliding time window is used to determine whether there is a pulse start position and / or pulse end position in the next time window, until the signal to be processed is traversed. For each time window, the initial state of the pulse flag corresponding to the current time window inherits the final state of the pulse flag of the previous time window.

7. A signal adaptive detection device, characterized in that, The device includes: The acquisition module is used to acquire the signal to be processed, and to segment the signal to be processed based on a preset unit signal width to obtain multiple segments of the signal to be processed. The autocorrelation value calculation module is used to set a time window, and for each segment of signal to be processed within the time window, perform delayed autocorrelation processing on the segment of signal to be processed to obtain the delayed autocorrelation value. The first signal type determination module is used to perform histogram statistics on the delay autocorrelation calculation values ​​of all segmented signals to be processed within the time window to determine whether the signal type within the time window is unique. The second signal type determination module is used to perform frequency domain analysis on the segmented signal to be processed within the time window if the signal type is unique within the time window, and to determine the signal type within the time window. The first position detection module is used to perform pulse detection on the sliding time window in combination with the signal detection threshold corresponding to the signal type distribution within the time window, and to determine whether there is a pulse start position and / or pulse end position of the signal to be processed. The second position detection module is used to slide the time window, count the pulse position detection results within each step of the sliding time window, determine the start position and / or end position of all pulses in the signal to be processed, and obtain the signal detection result.

8. A signal adaptive detection device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the signal adaptive detection method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the signal adaptive detection method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the signal adaptive detection method as described in any one of claims 1 to 6.

Images