Adaptive anti-interference composite sensing alarm method

Abstract

This invention provides an adaptive anti-interference composite sensing alarm method, belonging to the field of adaptive anti-interference technology. The invention performs modal decomposition of vibration signals using a vibration decomposition model and calculates vibration signal decomposition optimization indices. Based on these indices, the parameters of the identification model are optimized. Target features of modal components are acquired, and dynamic unfamiliar attributes within a time period formed by multiple window times are calculated using the sliding window method. Initial microwave transmission parameters are set to detect the area to be warned. Time-frequency domain analysis of the echo signal is performed, and the detection quality of the echo signal is calculated. The microwave transmission parameters are optimized using a microwave optimization model to obtain the optimal echo signal. A feature recognition model is established, and the output feature recognition data is used for composite warning. When an obstacle obstructs the microwave signal, the diffraction wave is automatically adjusted for detection.

Description

Technical Field

[0001] This invention relates to the field of adaptive anti-interference technology, specifically to a composite sensing alarm method with adaptive anti-interference capabilities. Background Technology

[0002] With the increasing demand for security, especially in critical areas such as bases, prisons, factories, and commercial buildings, the performance and reliability of intrusion detection systems are of paramount importance. Traditional perimeter security monitoring typically relies on a single sensing method, such as vibration sensors or microwave detectors, but these methods face significant limitations. For example, vibration sensors are prone to false alarms or missed detections when faced with complex external interference. Vibration detection is susceptible to interference from environmental noise (such as wind, rain, vehicle traffic, and small animal activity), resulting in a high false alarm rate. Furthermore, sensitivity decreases in complex terrain or when the surface medium changes, posing a risk of missed detections. While microwave detectors can operate in all weather conditions, their signals are easily affected by severe weather (rain, fog, snow), vegetation, and obstructions, leading to unstable detection quality. Moreover, microwave detectors cannot accurately detect targets behind obstructions, resulting in blind spots in monitoring. Meanwhile, traditional systems have fixed parameters and lack the ability to adaptively adjust to environmental changes and interference. They also lack the ability to effectively distinguish between real intrusions and complex environmental interference, making it difficult to meet the real-time monitoring requirements of low false alarms, low false alarms, and high reliability in high-security scenarios. These technical problems pose significant challenges to practical applications. Therefore, there is an urgent need for a multimodal intrusion detection system that can combine vibration and microwaves and adaptively optimize performance to reduce the impact of external interference and improve detection accuracy.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide an adaptive anti-interference composite sensing alarm method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An adaptive anti-interference composite sensing alarm method, comprising the following steps:

[0007] Step 1: Collect vibration signals from the ground in the area to be warned, perform modal decomposition on the vibration signals using a vibration decomposition model, calculate vibration signal decomposition optimization index based on the modal clarity, number, and power spectral entropy of the decomposed target, and optimize the parameters of the vibration decomposition model based on the vibration signal decomposition optimization index.

[0008] Step 2: Obtain the target features of each modal component by performing deep decomposition on each modal component. Using the sliding window method, assign window unfamiliar attributes to the target features according to the duration and frequency of each target feature within each window time, and calculate the dynamic unfamiliar attributes within the time period formed by multiple window times.

[0009] Step 3: Obtain environmental data of the area to be warned, set the initial parameters for microwave transmission to detect the area to be warned, perform time-frequency domain analysis on the echo signal, and calculate the detection quality of the echo signal;

[0010] Step 4: Based on the detection quality of the echo signal, real-time environmental data, and the parameter range of the microwave detection equipment, optimize the microwave transmission parameters using a microwave optimization model to obtain the best echo signal;

[0011] Step 5: Establish a feature recognition model, output feature recognition data respectively, and perform composite early warning based on feature recognition data. When an obstacle that blocks the microwave signal is detected by microwave detection, the diffraction wave is automatically adjusted to detect behind the obstacle and issue an early warning.

[0012] Furthermore, the vibration decomposition model is based on variational mode decomposition, which decomposes the signal into multiple mode functions with specific center frequencies and bandwidths;

[0013] Furthermore, the method for optimizing the parameters of the vibration decomposition model based on the vibration signal decomposition optimization index is as follows:

[0014] The key parameters required for variational mode decomposition are optimized using the Grey Wolf optimization algorithm, and the number of decomposition layers is dynamically adjusted. With penalty factor The parameters, specifically the calculation steps, include constructing the wolf pack position matrix, calculating the distance vector, updating the position, and after completing the number of iterations, selecting the optimization parameter that maximizes the vibration signal decomposition optimization index.

[0015] Furthermore, the calculation method for the vibration signal decomposition optimization index based on the modal clarity, number, and power spectral entropy of the decomposed target is as follows:

[0016] The method for calculating power spectral entropy is as follows:

[0017] ;

[0018] ;

[0019] in, For power spectral entropy, For frequency in Power at that location, Modal components Fourier transform operation;

[0020] The formula for calculating modal clarity is:

[0021]

[0022] in, For the first One effective modal component, for Clarity, for Hilbert transform;

[0023] The calculation method for the vibration signal decomposition optimization index is as follows:

[0024] ;

[0025] in, To optimize the index for vibration signal decomposition, This represents the number of effective modal components.

[0026] Furthermore, the method for calculating the unfamiliarity attribute of the window is as follows:

[0027] ;

[0028] The calculation method for dynamic unfamiliar attributes is as follows:

[0029] ;

[0030] in, For unfamiliar attributes within the window time. The duration sensitivity coefficient, Duration of target appearance The number of occurrences within the window time. The number of windows within the time period. For the first The unfamiliar properties of a window of time For forgetting adjustment factor, The current time window number.

[0031] Furthermore, the microwave transmission parameters include transmission power, center frequency, beam direction angle, and pulse width;

[0032] Environmental parameters include temperature, humidity, air pressure, and particulate matter concentration.

[0033] Furthermore, the microwave optimization model is based on a sequential quadratic programming algorithm. The specific calculation steps involve constructing an environmental state vector and a microwave detection parameter vector, respectively, and establishing a transmission parameter optimization model and constraints based on the detection quality of the echo signal. The parameter vector of the sub-optimization For the objective function It is approximately an optimizable form with quadratic terms.

[0034] Furthermore, the feature recognition model includes a vibration recognition model and a microwave recognition model. The feature recognition model is based on a convolutional neural network and specifically includes a convolutional layer, an activation layer, a pooling layer, and an output layer.

[0035] Feature recognition data includes;

[0036] Vibration target type and location;

[0037] Data on the type, location, and size of microwave-detected targets;

[0038] The vibration recognition model is trained by taking the modal functions of the decomposed vibration signal as input and the vibration signal type and location as output. It identifies the vibration target type and location by taking the modal functions of the decomposed vibration signal in real time as input and outputs the confidence level of the recognition result.

[0039] The microwave recognition model is trained by taking the time-frequency domain features of the echo signal as input and the type, location, and size data of the microwave detection target as output. The model is trained by taking the time-frequency domain features of the real-time detected echo signal as input, identifying the type, location, and size data of the microwave detection target, and outputting the confidence level of the recognition result.

[0040] Furthermore, the method for conducting composite early warning based on feature recognition data is as follows:

[0041] ;

[0042] in, The confidence level after fusion. , These represent the confidence levels of the vibration signal and microwave signal output by the identification model, respectively. These are the weights for the confidence levels of the vibration signal and the microwave signal, respectively. These are the coordinate locations of vibration-induced intrusion targets and microwave-identified intrusion targets of the same type, respectively. For location tolerance parameters, , These are the correction factors for the confidence levels of vibration signals and microwave signals, respectively.

[0043] when , , Immediately issue a warning;

[0044] in, To integrate early warning thresholds, The vibration detection early warning threshold, This is the microwave detection early warning threshold.

[0045] Furthermore, the method for automatically adjusting the diffraction wave to detect behind the obstruction and issue an early warning when an obstacle blocking the microwave signal is detected is as follows:

[0046] Based on the target type, location, and size data identified by microwave detection, and the location of the detection equipment, the area behind the obstruction is detected by modulating the diffraction wave parameters. The modulation steps for the diffraction wave parameters are as follows:

[0047] Formula for calculating wavelength:

[0048] ;

[0049] in, The wavelength of the diffracted wave is... The lateral dimension of the obstruction. The straight-line distance between the radar and the obstacle. The angle between the microwave beam and the surface of the obstacle. This is the diffraction optimization factor;

[0050] Diffraction deflection angle of the main lobe of the beam:

[0051] ;

[0052] in, It is the diffraction deflection angle. This is the azimuth angle of the obstacle's edge.

[0053] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention performs modal decomposition of vibration signals through a vibration decomposition model and calculates vibration signal decomposition optimization index. Based on the vibration signal decomposition optimization index, the parameters of the identification model are optimized. By obtaining the target features of modal components, the dynamic unfamiliar attributes within the time period formed by multiple window times are calculated using the sliding window method. The area to be warned is detected by setting the initial parameters of microwave transmission. The echo signal is analyzed in the time and frequency domain and the detection quality of the echo signal is calculated. The microwave transmission parameters are optimized through the microwave optimization model to obtain the best echo signal. A feature recognition model is established, and the feature recognition data is output separately and a composite warning is given. When an obstacle that blocks the microwave signal is detected by microwave detection, the diffraction wave is automatically adjusted for detection.

[0054] This invention significantly improves the reliability and intelligence of perimeter security by integrating vibration and microwave sensing with an adaptive anti-interference composite sensing alarm mechanism. First, optimized vibration signal decomposition and dynamic unfamiliar attribute analysis effectively eliminate interference signals caused by environmental noise, greatly reducing the false alarm rate of vibration detection. Second, by analyzing environmental data and echo quality in real time and dynamically optimizing microwave transmission parameters based on sequential quadratic programming, the stability and high quality of the detection signal under adverse weather conditions are ensured, reducing the risk of missed alarms. Third, a feature recognition model performs composite analysis of dual-modal data, improving the accuracy of intrusion detection. Finally, when microwaves encounter obstruction, the system automatically adjusts the area behind the diffraction wave detection blind zone, eliminating blind spots in traditional microwave detection. Ultimately, this solution achieves all-weather, high-precision, and robust real-time perimeter intrusion monitoring in complex and variable environments, significantly improving security protection effectiveness. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the overall method flow of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0057] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0058] Example:

[0059] Please see Figure 1 The present invention provides a technical solution:

[0060] An adaptive anti-interference composite sensing alarm method, comprising the following steps:

[0061] Step 1: Collect vibration signals from the ground in the area to be warned, perform modal decomposition on the vibration signals using a vibration decomposition model, calculate vibration signal decomposition optimization index based on the modal clarity, number, and power spectral entropy of the decomposed target, and optimize the parameters of the identification model based on the vibration signal decomposition optimization index.

[0062] Variational mode decomposition (VMD) is a signal processing method used to decompose a complex signal into several sub-signals (called mode functions) with specific frequency characteristics. Compared with traditional empirical mode decomposition (EMD), VMD has higher mathematical rigor and robustness, making it highly suitable for the analysis and early warning systems of non-stationary vibration signals.

[0063] The main advantage of using variational mode decomposition (VMD) to decompose vibration signals is that it can effectively break down complex vibration signals into multiple mode functions with specific center frequencies and bandwidths, facilitating individual analysis of different frequency components. VMD is an adaptive decomposition technique that automatically determines the optimal number of modes and frequency range based on the signal characteristics, making it particularly suitable for processing nonlinear and non-stationary vibration signals. Through this decomposition, target features in the vibration signal can be accurately extracted with minimal interference between modes, thus reducing false identifications caused by signal aliasing. Furthermore, the adaptability of VMD effectively addresses signal variations caused by different environmental changes, such as temperature and humidity, enabling the early warning system to operate stably in various complex environments. In summary, using VMD to decompose vibration signals improves the accuracy of subsequent feature extraction and target recognition, resulting in a more robust and interference-suppressing overall solution in complex environments, ultimately achieving more accurate and reliable intrusion detection and early warning.

[0064] In this embodiment, the vibration decomposition model is based on variational mode decomposition, which decomposes the signal into multiple mode functions with specific center frequencies and bandwidths. The specific calculation formula is as follows:

[0065] Variational mode decomposition:

[0066] ;

[0067] Constraint functions:

[0068] ;

[0069] in, For the first One modal function, For the first The center frequency of each modal function For time partial derivative operators, For time partial derivative operators, The imaginary unit, It is a complex exponential modulation term. For the collected vibration signals, The number of modes decomposition layers. It is a time variable;

[0070] The accuracy and convergence of the decomposition are guaranteed by introducing a penalty term and Lagrange multipliers:

[0071] ;

[0072] in, As a penalty factor, For Lagrange multipliers, It is a Lagrange function.

[0073] The Grey Wolf Optimization Algorithm (GFA) is a swarm intelligence optimization algorithm that simulates the hunting behavior of grey wolves. It can find the global optimum in a complex search space. In variational mode decomposition (VMD), key parameters such as the number of decomposition levels and the penalty factor directly affect the decomposition effect and the clarity of the mode decomposition. Optimizing these key parameters through the Grey Wolf Optimization Algorithm can make the VMD decomposition results more consistent with the characteristics of actual signals, thus improving the accuracy of mode decomposition. First, the Grey Wolf Optimization Algorithm can simulate the hunting strategy of grey wolves in nature, performing a global search in the search space, thereby avoiding the trap of local optima and ensuring that the VMD algorithm can obtain optimal decomposition results under various noise and interference conditions. Second, the Grey Wolf Optimization Algorithm has strong global search capabilities and can consider various signal characteristics during parameter adjustment, improving the model's adaptability and robustness to complex signals, which is crucial for the nonlinear and non-stationary characteristics of vibration signals. By dynamically adjusting key parameters in VMD, such as the number of decomposition levels and the penalty factor, not only can the decomposition quality of vibration signals be improved, but the recognition model's resistance to interference and noise can also be enhanced, thereby improving the accuracy and reliability of the system and playing an important role in vibration early warning systems.

[0074] In practical applications, vibration signals are often affected by noise, interference, and nonlinear factors. Optimizing these key parameters allows the system to better decompose meaningful modal components, effectively suppress the influence of noise and interference, and improve the clarity and discernibility of signal decomposition. In this way, the system can not only extract target modes more accurately but also enhance its robustness to various interferences, thereby improving the performance and reliability of the entire recognition model.

[0075] In this embodiment, the method for optimizing the parameters of the identification model based on the vibration signal decomposition optimization index is as follows:

[0076] The key parameters required for variational mode decomposition are optimized using the Grey Wolf optimization algorithm, and the number of decomposition layers is dynamically adjusted. With penalty factor The parameters, and the specific calculation steps are as follows:

[0077] Constructing the wolf pack position matrix :

[0078] ;

[0079] in, For the first The number of decomposition layers of a lone wolf. For the first The punishment factor of a lone wolf The size of a wolf pack;

[0080] Calculate the distance vector:

[0081] ;

[0082] Location update:

[0083] ;

[0084] in, The distance between the gray wolf and its prey. For the prey The position at time, i.e., the position of the optimal parameters. In order to be in The location of the wolf pack at all times. For random disturbance factors, For coefficient vectors;

[0085] After completing the required number of iterations, the optimal parameter with the highest vibration signal decomposition optimization index is selected.

[0086] By calculating vibration signal decomposition optimization indices based on the modal clarity, number of modal components, and power spectral entropy of the decomposed target, the quality and accuracy of vibration signal decomposition can be effectively improved, thereby enhancing the system's adaptive anti-interference capability. Modal clarity reflects the independence and discernibility of different modal components; higher clarity indicates more accurate signal decomposition and reduces the impact of noise and interference on the signal. The number of modal components helps assess the sufficiency of signal decomposition, ensuring that important frequency components are not missed during the decomposition process. Power spectral entropy characterizes the complexity of the signal; signals with lower entropy values ​​generally indicate simpler structures, which helps improve the stability of the recognition model in complex environments. Combining these three indices not only optimizes the signal decomposition process but also adaptively adjusts parameters to adapt to different interference and noise conditions, improving the model's robustness. In dynamic environments, this calculation of vibration signal-based optimization indices enables the system to adapt more accurately to various changes, improves resistance to interference, and ensures efficient and stable target recognition tasks under various complex situations.

[0087] In this embodiment, the method for calculating the vibration signal decomposition optimization index based on the modal clarity, number, and power spectral entropy of the decomposed target is as follows:

[0088] The method for calculating power spectral entropy is as follows:

[0089] ;

[0090] ;

[0091] in, For power spectral entropy, For frequency in Power at that location, Modal components Fourier transform operation;

[0092] The formula for calculating modal clarity is:

[0093] ;

[0094] in, For the first One effective modal component, for Clarity, for Hilbert transform;

[0095] The calculation method for the vibration signal decomposition optimization index is as follows:

[0096] ;

[0097] in, To optimize the index for vibration signal decomposition, This represents the number of effective modal components.

[0098] Step 2: Obtain the target features of each modal component by performing deep decomposition on each modal component. Using the sliding window method, assign window unfamiliar attributes to the target features according to the duration and frequency of each target feature within each window time, and calculate the dynamic unfamiliar attributes within the time period formed by multiple window times.

[0099] By employing the sliding window method, a window attribute is assigned to the target based on the duration and frequency of each target feature within each window period. This helps to dynamically capture the temporal characteristics and patterns of target behavior, thereby enhancing the system's ability to identify abnormal signals and interference, and improving its adaptive anti-interference performance. Duration reflects the stability of a feature over time, while the frequency of occurrence reflects its repeatability and regularity. The combination of these two factors effectively distinguishes between real targets and occasional interference signals. The sliding window mechanism enables dynamic updates of local features along the time axis, allowing the system to maintain high-precision identification even when facing non-stationary interference or sudden disturbances. This facilitates rapid response to environmental changes and adaptive parameter adjustment, effectively improving the system's robustness and reliability, and maintaining stable tracking and accurate perception of targets in complex environments.

[0100] By assigning a forgetting coefficient to each window and calculating dynamic unfamiliar attributes across multiple window time periods, the system can effectively address the impact of factors such as rainfall and wind on vibration detection equipment. Particularly effective against interfering signals, the system can gradually reduce their influence on system judgment over a period of time, achieving adaptive anti-interference. For example, vibration signals generated by continuous rainfall may initially affect system judgment. The system analyzes information such as characteristic frequencies and amplitudes within the window. If these signals are prolonged and recurring, the system gradually assigns a higher forgetting coefficient to them, reducing the weight of historical interference signals. Ultimately, through re-evaluation of the target characteristics, vibration signals generated by non-hazardous factors such as rainfall or wind are ignored, thus eliminating false alarms. Furthermore, in scenarios involving wind blowing through railings, the system can identify and gradually eliminate prolonged, non-hazardous vibration features through dynamic unfamiliar attributes, ensuring that warnings are not issued incorrectly due to environmental interference. Therefore, the dynamic adjustment of the forgetting coefficient allows the system to flexibly adapt to environmental changes and effectively filter out non-hazardous interference signals, enhancing the stability and accuracy of vibration detection equipment in complex environments.

[0101] In this embodiment, the method for calculating the window unfamiliarity attribute is as follows:

[0102] ;

[0103] The calculation method for dynamic unfamiliar attributes is as follows:

[0104] ;

[0105] in, For unfamiliar attributes within the window time. The duration sensitivity coefficient, Duration of target appearance The number of occurrences within the window time. The number of windows within the time period. For the first The unfamiliar properties of a window of time For forgetting adjustment factor, The current time window number.

[0106] Step 3: Obtain environmental data of the area to be warned, set the initial parameters for microwave transmission to detect the area to be warned, perform time-frequency domain analysis on the echo signal, and calculate the detection quality of the echo signal.

[0107] By performing time-frequency domain analysis on echo signals using wavelet transform and calculating the detection quality, the time and frequency domain features of the signal can be effectively extracted, thereby improving the system's adaptive anti-interference capability. Wavelet transform possesses excellent time-frequency localization characteristics, balancing high time resolution and high frequency resolution, which is particularly important for detecting interference signals with rapidly changing frequencies or short durations. For example, in practical applications, changes in environmental temperature, humidity, or air pressure can alter the propagation characteristics of microwave signals, affecting the quality of the echo signal. In such cases, wavelet transform can effectively decompose the interference signal into the target signal. During time-frequency domain analysis, the system can identify abnormal changes in the echo signal in real time, such as high-frequency noise appearing within a short period or low-frequency interference caused by weather changes. Through this time-frequency domain decomposition, the quality of the echo signal can be accurately assessed, and components within the echo signal can be identified as interference signals and valuable target signals, thus avoiding misjudging environmental interference as the target signal. Furthermore, combining microwave transmission parameters (such as power, frequency, beam angle, etc.) with environmental state parameters (such as temperature, humidity, etc.) and using time-frequency analysis through wavelet transform can help further improve the robustness of the system. This allows for adjustments to the detection strategy based on real-time environmental changes, ensuring the stability and accuracy of the system in complex and dynamic environments.

[0108] The method for performing time-frequency domain analysis on echo signals is as follows:

[0109] ;

[0110] in, The time-frequency domain characteristics of the echo signal, As a scale factor, The translation factor is... For wavelet basis functions, It is a complex conjugate.

[0111] In this embodiment, the microwave transmission parameters include transmission power, center frequency, beam direction angle, and pulse width;

[0112] Environmental parameters include temperature, humidity, air pressure, and particulate matter concentration;

[0113] The method for calculating the detection quality of the echo signal is as follows:

[0114] ;

[0115] ;

[0116] in, For the detection quality of the echo signal, For frequency band The corresponding scale, For time window, For frequency band The number of sampling points, For frequency band energy, For the target feature frequency band set, For frequency band The weight, This is the noise suppression coefficient. Background noise level, This represents noise power.

[0117] Step 4: Based on the detection quality of the echo signal, real-time environmental data, and the parameter range of the microwave detection equipment, optimize the microwave transmission parameters using a microwave optimization model to obtain the best echo signal.

[0118] Sequential quadratic programming (SQP) is an iterative algorithm for solving nonlinear constrained optimization problems. Its basic idea is to approximate the original nonlinear problem as a quadratic programming subproblem in each iteration. By continuously solving this subproblem and updating the variables, the optimal solution is gradually approximated. SQP is efficient and accurate, especially suitable for complex optimization problems where both the objective function and constraints are nonlinear. In microwave transmission parameter optimization, multiple coupled variables (such as frequency, power, and pulse width) are involved, and the optimization is affected by various constraints such as hardware limitations, signal propagation laws, and interference background. The optimization objective is the detection quality of the echo signal. This type of problem is inherently nonlinear and constrained. Therefore, the SQP algorithm can effectively handle these complex coupling relationships and physical constraints, achieving fine-tuning of microwave parameters, thereby obtaining the best echo effect and improving the performance and anti-interference capability of the detection system.

[0119] Constructing a microwave optimization model and obtaining the optimal echo signal using a sequential quadratic programming algorithm aims to balance various constraints and objective functions, and is particularly suitable for complex optimization problems with constraints. In microwave detection systems, the quality of the echo signal is affected by microwave transmission parameters (such as power, frequency, and pulse width) and environmental factors (such as temperature, humidity, and air pressure). The sequential quadratic programming algorithm allows for precise adjustment of microwave transmission parameters based on real-time echo signal quality and environmental data, optimizing detection performance. For example, when strong interference signals (such as meteorological factors or stray noise) occur in the environment, the sequential quadratic programming algorithm can dynamically adjust microwave transmission power, frequency, and other parameters while considering system constraints, optimizing echo signal quality, improving detection quality, and avoiding the effects of interference. Specifically, in practical operation, if temperature changes cause variations in microwave propagation characteristics, the sequential quadratic programming algorithm can automatically calculate the most suitable frequency and power combination to achieve optimal echo signal quality, thereby ensuring the system maintains stable detection capabilities even in harsh environments. This method not only improves the adaptability of the detection system in complex environments but also enhances its anti-interference capabilities, ensuring that it can effectively distinguish between target signals and interference signals, thereby improving the accuracy and reliability of detection.

[0120] In this embodiment, the microwave optimization model is based on a sequential quadratic programming algorithm. The specific calculation steps are as follows: construct the environmental state vector and the microwave detection parameter vector respectively; and establish a transmission parameter optimization model and constraints based on the detection quality of the echo signal. The parameter vector of the sub-optimization For the objective function It is approximated as an optimizable form with quadratic terms. The specific calculation steps are as follows:

[0121] Constructing the environment state vector:

[0122] ;

[0123] in, Let the environment state vector be... For temperature, For humidity, For air pressure, This refers to the particulate matter concentration.

[0124] Construct the parameter vector for microwave detection:

[0125] ;

[0126] in, For parameter vectors, For transmission power, For the center frequency, The beam direction angle, The pulse width;

[0127] Construct a launch parameter optimization model:

[0128] ;

[0129] Constraints:

[0130] ;

[0131] in, These are the lower and upper limits of the transmission power, respectively. This represents the range of center frequencies. The range of beam direction angle, The range of pulse width;

[0132] Construct a quadratic programming subproblem:

[0133] Put the first The parameter vector of the sub-optimization For the objective function It can be approximated as an optimizable form with quadratic terms:

[0134] ;

[0135] in, Increment of the parameter vector For the first Curvature of the objective function in the next iteration For the first The gradient of the objective function in the next iteration This is a transpose operation;

[0136] Constrained linearization:

[0137] ;

[0138] in, For the first The parameter vector for the next iteration. Let be the constraint function vector, representing the parameter vector. All constraint values ​​at the location, To constrain the Jacobian matrix, This is the sequence number of the current optimization step.

[0139] Step 5: Establish a feature recognition model, output feature recognition data respectively, and perform composite early warning based on feature recognition data. When an obstacle that blocks the microwave signal is detected by microwave detection, the diffraction wave is automatically adjusted to detect behind the obstacle and issue an early warning.

[0140] Convolutional Neural Networks (CNNs) are widely used in feature recognition, especially for processing spatiotemporal features in signals such as vibration and microwave signals. For vibration recognition models, CNNs can automatically learn discriminative features from signals, particularly through decomposed mode functions. These mode functions effectively extract different frequency components of the vibration signal, and the time-frequency domain features obtained through wavelet transform further improve the signal's resolution and robustness. In microwave recognition models, CNNs can process the time-frequency domain features of the echo signal obtained after wavelet transform, accurately distinguishing microwave detection targets from other background noise or interference signals. By establishing these two models separately, the system can perform specialized processing and recognition for different types of signals, thereby improving the accuracy and reliability of target recognition, especially in complex environments or when facing signal interference. For example, in practical applications, when the detection system is affected by obstructions, the vibration recognition model can identify the physical changes caused by vibration, while the microwave recognition model can accurately detect target signals behind the obstruction. It can then adaptively adjust the diffraction wave for more precise target detection and recognition, and issue warnings based on the recognition results. In this way, multimodal data fusion can effectively improve the system's anti-interference ability and ensure continuous and stable operation under complex conditions.

[0141] In this embodiment, the feature recognition model includes a vibration recognition model and a microwave recognition model. The feature recognition model is based on a convolutional neural network and specifically includes a convolutional layer, an activation layer, a pooling layer, and an output layer.

[0142] Feature recognition data includes;

[0143] Vibration target type and location;

[0144] Data on the type, location, and size of microwave-detected targets;

[0145] The vibration recognition model is trained by taking the modal functions of the decomposed vibration signal as input and the vibration signal type and location as output. It identifies the vibration target type and location by taking the modal functions of the decomposed vibration signal in real time as input and outputs the confidence level of the recognition result.

[0146] The microwave recognition model is trained by taking the time-frequency domain features of the echo signal as input and the type, location, and size data of the microwave detection target as output. The model is trained by taking the time-frequency domain features of the real-time detected echo signal as input, identifying the type, location, and size data of the microwave detection target, and outputting the confidence level of the recognition result.

[0147] The convolutional neural network specifically includes a convolutional layer, an activation layer, a pooling layer, and an output layer;

[0148] Convolutional layer;

[0149] ;

[0150] in, The modal functions and time-frequency domain characteristics of the echo signal are decomposed from the input vibration signal. For convolution kernel, To output the coordinates of the data matrix, The convolution kernel in the th row and number The value of the column;

[0151] The formula for calculating the activation layer is:

[0152] ;

[0153] in, The coordinates of the output matrix of the convolutional layer;

[0154] The formula for calculating the pooling layer is:

[0155] ;

[0156] in, This is for max-pooling output;

[0157] The formula for calculating the output layer is:

[0158] ;

[0159] in, This is the output of the pooling layer. The result of feature recognition data identification. As weight, This is the bias parameter.

[0160] By combining the location and confidence levels of vibration-induced intrusion targets and microwave-detected targets for composite early warning, complementary enhancement of multi-source information can be achieved, thereby improving the accuracy and robustness of target identification and intrusion detection in complex environments. Vibration detection is primarily sensitive to disturbances on the ground or structures, accurately identifying the location of physical contact behaviors such as touching and climbing, but it is easily affected by environmental noise such as wind and mechanical vibration. Microwave detection, on the other hand, is suitable for identifying the movement trajectory of targets in the air or at long distances, especially penetrating a certain degree of obstruction, but it is easily affected by factors such as rain, fog, and electromagnetic interference. By extracting the location and confidence information from the outputs of the two models separately, they can be mutually verified and supplemented in practical applications, improving the reliability of the overall identification results. Especially when the signal is incomplete or a single sensor is interfered with, the system can make fault-tolerant judgments and issue early warnings based on information provided by another signal source, thereby improving adaptive anti-interference capabilities.

[0161] For example, in a boundary protection scenario, if an obstruction on one side severely attenuates the microwave signal, and the system cannot acquire a clear image even after automatically adjusting the diffraction wave, the vibration model detects obvious vibration characteristics on the ground in that area. Combining this with the output location coordinates and a high recognition confidence level, the system uses data fusion to confirm the presence of intrusion risk and promptly triggers a composite early warning. Conversely, if strong winds cause a vibration sensor to falsely trigger, and the microwave detection model fails to identify a moving target or has a low recognition confidence level at the same location, the system can automatically reduce the signal weight to avoid false alarms. This location and confidence level fusion mechanism based on multi-source identification results not only enhances anti-interference capabilities but also makes the early warning results more stable and reliable.

[0162] In this embodiment, the method for performing composite early warning based on feature recognition data is as follows:

[0163] ;

[0164] in, The confidence level after fusion. , These represent the confidence levels of the vibration signal and microwave signal output by the identification model, respectively. , These are the weights for the confidence levels of the vibration signal and the microwave signal, respectively. These are the coordinate locations of vibration-induced intrusion targets and microwave-identified intrusion targets of the same type, respectively. For location tolerance parameters, , These are the correction factors for the confidence levels of vibration signals and microwave signals, respectively.

[0165] when , , Immediately issue a warning;

[0166] in, To integrate early warning thresholds, The vibration detection early warning threshold, This is the microwave detection early warning threshold.

[0167] In this embodiment, the method for automatically adjusting the diffraction wave to detect behind the obstruction and provide an early warning when an obstacle blocking the microwave signal is detected is as follows:

[0168] Based on the target type, location, and size data identified by microwave detection, and the location of the detection equipment, the area behind the obstruction is detected by modulating the diffraction wave parameters. The modulation steps for the diffraction wave parameters are as follows:

[0169] Formula for calculating wavelength:

[0170] ;

[0171] Center frequency:

[0172] ;

[0173] How bandwidth is calculated:

[0174] ;

[0175] in, The wavelength of the diffracted wave is... The lateral dimension of the obstruction. The straight-line distance between the radar and the obstacle. The angle between the microwave beam and the surface of the obstacle. For diffraction optimization factor, For the center frequency, bandwidth, The speed of light;

[0176] Diffraction deflection angle of the main lobe of the beam:

[0177] ;

[0178] in, It is the diffraction deflection angle. This is the azimuth angle of the obstacle's edge.

[0179] When microwave detectors encounter obstructions, they automatically adjust the diffraction wave for detection. This is to overcome signal attenuation or complete loss caused by obstacles, ensuring that targets in the detection area can still be detected. Microwave signals typically undergo reflection, refraction, or diffraction when encountering obstacles; diffraction refers to the way signals propagate around obstacles. In many applications, obstacles may completely block microwave signals, preventing targets from being detected via conventional straight-line propagation paths. By adjusting the diffraction wave, the microwave detector can adjust the signal propagation path, allowing the signal to bypass obstacles and accurately detect targets behind them, even when obstructed. This method effectively solves the blind zone problem that traditional microwave detectors may encounter when facing obstacles, improving the system's continuous monitoring capabilities, and maintaining high detection accuracy, especially in complex environments.

[0180] For example, in a city security system, if a microwave sensor is obstructed by a building or other tall object, the system will automatically adjust the diffraction angle and propagation mode of the microwave signal to bypass the building's obstruction and detect targets behind it. If a person or object attempts to enter the protected area, the diffracted wave will enable the microwave signal to detect the target's presence and trigger an alarm. Through this adaptive adjustment, the system can continuously monitor and reduce detection blind spots caused by obstructions when facing various natural or man-made obstacles, thereby effectively improving its defense capabilities and ensuring the stable operation of the security system in complex environments.

[0181] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0182] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0183] The units described as decomposed components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0184] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes 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.

Claims

1. A composite sensing alarm method with adaptive anti-interference capabilities, characterized in that, The specific steps include: Step 1: Collect vibration signals from the ground in the area to be warned, perform modal decomposition on the vibration signals using a vibration decomposition model, calculate vibration signal decomposition optimization index based on the modal clarity, number and power spectral entropy of the decomposed target, and optimize the parameters of the vibration decomposition model based on the vibration signal decomposition optimization index. Step 2: Obtain the target features of each modal component by performing deep decomposition on each modal component. Using the sliding window method, assign window unfamiliar attributes to the target features according to the duration and frequency of each target feature within each window time, and calculate the dynamic unfamiliar attributes within the time period formed by multiple window times. Step 3: Obtain environmental data of the area to be warned, set the initial parameters for microwave transmission to detect the area to be warned, perform time-frequency domain analysis on the echo signal, and calculate the detection quality of the echo signal; Step 4: Based on the detection quality of the echo signal, real-time environmental data, and the parameter range of the microwave detection equipment, optimize the microwave transmission parameters using a microwave optimization model to obtain the best echo signal; Step 5: Establish a feature recognition model, output feature recognition data respectively, and perform composite early warning based on feature recognition data. When an obstacle that blocks the microwave signal is detected by microwave detection, the diffraction wave is automatically adjusted to detect behind the obstacle and issue an early warning.

2. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The vibration decomposition model is based on variational mode decomposition, which decomposes the signal into multiple mode functions with specific center frequencies and bandwidths.

3. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The method for optimizing the parameters of the vibration decomposition model based on the vibration signal decomposition optimization index is as follows: The key parameters required for variational mode decomposition are optimized using the Grey Wolf optimization algorithm, and the number of decomposition layers is dynamically adjusted. With penalty factor The parameters, specifically the calculation steps, include constructing the wolf pack position matrix, calculating the distance vector, updating the position, and after completing the number of iterations, selecting the optimization parameter that maximizes the vibration signal decomposition optimization index.

4. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The calculation method for the vibration signal decomposition optimization index based on the modal clarity, number, and power spectral entropy of the decomposed target is as follows: The method for calculating power spectral entropy is as follows: ； ； in, For power spectral entropy, For frequency in Power at that location, Modal components Fourier transform operation; The formula for calculating modal clarity is: ； in, For the first One effective modal component, for Clarity, for Hilbert transform; The calculation method for the vibration signal decomposition optimization index is as follows: ； in, To optimize the index for vibration signal decomposition, This represents the number of effective modal components.

5. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The method for calculating the unfamiliarity attribute of the window is as follows: ； The calculation method for dynamic unfamiliar attributes is as follows: ； in, For unfamiliar attributes within the window time. The duration sensitivity coefficient, Duration of target appearance The number of occurrences within the window time. The number of windows within the time period. For the first The unfamiliar properties of a window of time For forgetting adjustment factor, The current time window number.

6. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The microwave transmission parameters include transmission power, center frequency, beam direction angle, and pulse width; Environmental parameters include temperature, humidity, air pressure, and particulate matter concentration.

7. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The microwave optimization model is based on a sequential quadratic programming algorithm. The specific calculation steps are as follows: construct the environmental state vector and the microwave detection parameter vector respectively; and establish a transmission parameter optimization model and constraints based on the detection quality of the echo signal. The parameter vector of the sub-optimization For the objective function It is approximately an optimizable form with quadratic terms.

8. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The feature recognition model includes a vibration recognition model and a microwave recognition model. The feature recognition model is based on a convolutional neural network and specifically includes a convolutional layer, an activation layer, a pooling layer, and an output layer. Feature recognition data includes; Vibration target type and location; Data on the type, location, and size of microwave-detected targets; The vibration recognition model is trained by taking the modal functions of the decomposed vibration signal as input and the vibration signal type and location as output. It identifies the vibration target type and location by taking the modal functions of the decomposed vibration signal in real time as input and outputs the confidence level of the recognition result. The microwave recognition model is trained by taking the time-frequency domain features of the echo signal as input and the type, location, and size data of the microwave detection target as output. The model is trained by taking the time-frequency domain features of the real-time detected echo signal as input, identifying the type, location, and size data of the microwave detection target, and outputting the confidence level of the recognition result.

9. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The method for conducting composite early warning based on feature recognition data is as follows: ； in, The confidence level after fusion. , These represent the confidence levels of the vibration signal and microwave signal output by the identification model, respectively. , These are the weights for the confidence levels of the vibration signal and the microwave signal, respectively. These are the coordinate locations of vibration-induced intrusion targets and microwave-identified intrusion targets of the same type, respectively. For location tolerance parameters, , These are the correction factors for the confidence levels of vibration signals and microwave signals, respectively. when , , Immediately issue a warning; in, To integrate early warning thresholds, The vibration detection early warning threshold, This is the microwave detection early warning threshold.

10. The adaptive anti-interference composite sensing alarm method according to claim 1, characterized in that: The method for automatically adjusting the diffraction wave to detect behind the obstruction and issue an early warning when an obstacle blocking the microwave signal is detected is as follows: Based on the target type, location, and size data identified by microwave detection, and the location of the detection equipment, the area behind the obstruction is detected by modulating the diffraction wave parameters. The modulation steps for the diffraction wave parameters are as follows: Formula for calculating wavelength: ； in, The wavelength of the diffracted wave is... The lateral dimension of the obstruction. The straight-line distance between the radar and the obstacle. The angle between the microwave beam and the surface of the obstacle. This is the diffraction optimization factor; Diffraction deflection angle of the main lobe of the beam: ； in, It is the diffraction deflection angle. This is the azimuth angle of the obstacle's edge.

Images