A weak signal processing method for power amplifier control

Abstract

The present invention discloses a weak signal processing method for power amplifier control, which belongs to the technical field of power amplifiers, and specifically comprises: monitoring the vibration of a building in real time by a multi-axis vibration sensor, and collecting weak signals generated by the vibration in real time by using wireless transmission technology; the collected signals are amplified and processed by a power amplifier control algorithm, and denoised by an adaptive denoising algorithm to obtain a clear denoised signal; spectrum and time domain analysis are performed on the denoised signal, and the vibration trend and risk of the building are predicted by combining multi-source data fusion and deep learning models; according to the prediction results, the system formulates a power amplifier control strategy and automatically adjusts the power amplifier parameters to achieve vibration suppression; when the vibration exceeds a preset threshold or meets the warning conditions, the system automatically triggers the warning mechanism, sends warning information and starts emergency response measures to ensure the safety of the building.

Description

Technical Field

[0001] The invention belongs to the technical field of power amplifiers, and in particular relates to a weak signal processing method for controlling a power amplifier. Background Art

[0002] With the acceleration of urbanization, the number of high-rise buildings and large building structures is increasing, and building safety has become the focus of social attention. As one of the important factors affecting building safety, the monitoring and control of building vibration is particularly important. However, in the process of building vibration monitoring, due to environmental noise and signal attenuation factors, the monitored signal is often weak and difficult to be directly used for power amplifier control. However, the power amplifier plays a vital role in wireless communication systems. It is responsible for enhancing the power of the signal before the signal is transmitted to ensure that the signal can cover a sufficient distance during the transmission process and overcome the attenuation and interference in the environment. Traditional power amplifier control methods mainly focus on signal power amplification and linearity optimization, but often ignore the special needs of weak signal processing; weak signals, that is, signals with low power and small amplitude, are easily affected by noise and interference during transmission, resulting in signal quality degradation or even failure to be correctly received.

[0003] For example, the patent with authorization announcement number CN211720527U discloses a power amplifier with shielding function, including: a supporting base, the supporting base is square in shape, and a PCB board is fixedly installed inside the supporting base by fixing screw threads, a circuit is soldered to the wiring terminal of the PCB board, one end of the circuit is soldered to a power supply wire, one end of the power supply wire is electrically connected to an external wiring terminal, a signal processing chip, a filter control chip and a wifi chip are electrically soldered to the surface of the PCB board, and a signal regulator is fixedly installed in the middle of the top of the supporting base. The utility model facilitates the processing of the external searched signal through the filter through the supporting base, signal regulator and filter, and shields the weaker wifi signal, so as to facilitate filtering of external signals during daily use and improve the signal filtering strength.

[0004] The above existing technologies all have the following problems: 1) The power amplifier with shielding function mainly focuses on filtering and enhancing WiFi signals, but does not involve real-time monitoring and early warning functions of building vibrations, making it appear to be single-function in complex environments that require simultaneous processing of vibration monitoring and signal filtering; 2) Lack of intelligence and accuracy in processing complex signals; 3) Lack of sufficient response capabilities when faced with emergencies. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a weak signal processing method for power amplifier control, which monitors the vibration of the building in real time through a multi-axis vibration sensor, and uses wireless transmission technology to collect the weak signal generated by the vibration in real time; the collected signal is amplified and processed by the power amplifier control algorithm, and denoised by an adaptive denoising algorithm to obtain a clear denoised signal; the denoised signal is subjected to spectral and time domain analysis, combined with multi-source data fusion and deep learning models to predict the vibration trend and risk of the building; according to the prediction results, the system formulates a power amplifier control strategy and automatically adjusts the power amplifier parameters to achieve vibration suppression; when the vibration exceeds the preset threshold or meets the warning conditions, the system automatically triggers the warning mechanism, sends a warning message and initiates emergency response measures to ensure the safety of the building.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A weak signal processing method for power amplifier control, comprising:

[0008] Step S1: Use a multi-axis vibration sensor to monitor the building in real time, connect the sensor to a data acquisition device, and collect weak signals generated by vibration in real time through wireless transmission;

[0009] Step S2: setting the amplification factor, amplifying the collected weak vibration signal through the power amplifier control algorithm, and denoising the amplified signal using the adaptive denoising algorithm to obtain a denoised signal;

[0010] Step S3: Perform spectrum analysis and time domain analysis on the denoised signal to extract characteristic information of the signal, combine the results of multi-source data fusion, predict the vibration trend and risk of the building through a deep learning model, formulate a power amplifier control strategy based on the prediction results and the extracted signal characteristic information, apply the formulated power amplifier control strategy to the actual power amplifier control system, and output the processed signal data to a display device for display;

[0011] Step S4: Set the vibration threshold and warning conditions. When the vibration of the building exceeds the vibration threshold or meets the warning conditions, the warning mechanism is automatically triggered to send warning information to the user. At the same time, emergency response measures are automatically initiated according to the warning level and the actual situation of the building.

[0012] Specifically, the specific steps of step S3 include:

[0013] S3.1: Use Fourier transform to perform spectrum analysis on the denoised signal to obtain the signal spectrum diagram, and extract frequency domain features from the spectrum diagram. At the same time, observe the waveform, peak value, and period of the denoised signal in the time domain to extract time domain features;

[0014] S3.2: Introduce multi-dimensional environmental data, fuse frequency domain features and time domain features with multi-dimensional environmental data to obtain fused data, pass the fused data as input to the input layer of the neural network model, and pass the input data to the hidden layer starting from the input layer. At the same time, in each layer, calculate the output of each neuron according to the weight and bias term of the current layer;

[0015] S3.3: Use the activation function to perform nonlinear transformation on the output of the neuron, and pass the fused data after nonlinear transformation to the next layer until it reaches the output layer. At the same time, at the output layer, the vibration trend and risk results of the building predicted by the neural network are obtained.

[0016] Specifically, the specific steps of step S3 also include:

[0017] S3.4: Based on the prediction results of the neural network model and the extracted frequency domain characteristics and time domain characteristics information, analyze whether the power amplifier is working in the saturation region and whether the performance index requirements are met, and formulate the power amplifier control strategy through the control algorithm;

[0018] S3.5: converting the formulated power amplifier control strategy into execution instructions and applying it to the actual power amplifier control system to execute the control strategy by adjusting the parameters of the power amplifier;

[0019] S3.6: Output the processed spectrum diagram, time domain waveform, and prediction results to a display device, and display the vibration conditions and control effects of the building in an animated form.

[0020] Specifically, the calculation formula for the neuron output result in S3.2 is:

[0021]

[0022] in, represents the output of the i-th neuron in the a-th layer, A da (·) represents the adaptive activation function, represents the weight from the jth neuron in the (a-1)th layer to the ith neuron in the ath layer, A te (·) represents the function of attention weight, represents the output of the jth neuron in the (a-1)th layer, represents the bias term of the i-th neuron in the a-th layer, Represents the output of the i-th neuron in the (a-1)-th layer.

[0023] Specifically, the specific steps of formulating the power amplifier control strategy in S3.4 include:

[0024] S3.41: Setting a control target of the power amplifier according to the performance analysis result;

[0025] S3.42: Based on the prediction results of the neural network model and the real-time feedback signal, use the predictive control algorithm to adjust the control parameters of the power amplifier, and evaluate the performance of the predictive control algorithm based on the adjustment results;

[0026] S3.43: Use the gradient descent optimization control algorithm to globally optimize the control parameters and find the optimal parameter combination;

[0027] S3.44: Integrate the prediction model, control algorithm, and optimization algorithm into a unified control system for comprehensive monitoring and evaluation, and use the evaluation results as input to the control algorithm and adjustment strategy to achieve closed-loop feedback control.

[0028] Specifically, the formula form of the control target in S3.41 is:

[0029]

[0030] Among them, P 0，max Indicates the maximum output power, U0 indicates the effective value of the output voltage, I0 indicates the effective value of the output current, min(THD k ) represents the minimum total harmonic distortion, U k represents the voltage amplitude of the kth harmonic, ζ represents the harmonic frequency, U1 is the fundamental voltage amplitude, η max represents the maximum energy conversion efficiency of the power amplifier, P1 represents the signal power obtained by the load, V cc Indicates the power supply voltage, I avg represents the average current flowing through the power supply, ΔT represents the parameter fluctuation range, T max 、T min They represent the maximum and minimum parameter values ​​during the working process, BW represents bandwidth, and f H 、f L They represent the upper cutoff frequency and the lower cutoff frequency respectively, SNR represents the signal-to-noise ratio, S represents the signal power, and N represents the noise power.

[0031] Specifically, the specific formula for evaluating the performance of the predictive control algorithm in S3.42 is:

[0032]

[0033] Where Q represents the performance indicator of the predictive control algorithm, [m, n] represents the time range for evaluating the tracking error, and w q represents the weight of each prediction time step, t represents time, x e (t+q) represents the expected response at time (t+q), q represents the time step, and x mol(t+q) represents the response of the neural network model at time (t+q), β1 and β2 represent the weight factors of the control signal increment and smoothness, respectively, and n u represents the time range for evaluating the control increment, u′(t+q-1), u′(t+q-2), and u′(t+q) represent the control signals obtained by the predictive control algorithm at (t+q-1), (t+q-2), and (t+q), respectively, and λ l Represents the weight factor of the constraint condition, g l (u′) represents the constraint condition of the control signal u′, and r represents the number of the constraint conditions.

[0034] Specifically, an electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a weak signal processing method for power amplifier control when executing the computer program.

[0035] Specifically, a computer-readable storage medium stores computer instructions, and when the computer instructions are executed, steps of a weak signal processing method for power amplifier control are executed.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. The present invention proposes a weak signal processing method for power amplifier control. Through real-time monitoring of buildings by multi-axis vibration sensors, combined with wireless transmission technology, the weak vibration signals generated by buildings can be collected in real time. This real-time performance ensures that the system can quickly capture any possible vibration events and improves the timeliness of early warning. At the same time, after power amplification and adaptive denoising processing, the denoised signal obtained is more accurate; the intelligent analysis method is used to more accurately identify the characteristics and laws of vibration, thereby improving the accuracy of prediction.

[0038] 2. The present invention proposes a weak signal processing method for power amplifier control. The system can formulate a power amplifier control strategy based on the prediction results and the extracted signal feature information, and adjust the parameters of the power amplifier in real time. This adaptability and flexibility enables the system to respond accordingly according to different vibration conditions, thereby improving the system's response speed and coping capabilities. The early warning and emergency response settings reduce the losses caused by potential risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A schematic diagram of a weak signal processing method for power amplifier control according to the present invention;

[0040] Figure 2 A flowchart of a weak signal processing method for power amplifier control according to the present invention;

[0041] Figure 3 A flowchart of a denoising signal acquisition method for weak signal processing used for power amplifier control according to the present invention;

[0042] Figure 4 The present invention is a flowchart of a power amplifier control strategy implementation method for a weak signal processing method for power amplifier control. DETAILED DESCRIPTION

[0043] Example 1

[0044] See also Figure 1-Figure 2 , an embodiment of the present invention provides: a weak signal processing method for power amplifier control, comprising the following steps:

[0045] Step S1: Use a multi-axis vibration sensor to monitor the building in real time, connect the sensor to a data acquisition device, and collect weak signals generated by vibration in real time through wireless transmission. The present invention ensures that vibration signals in all directions can be captured by deploying a multi-axis vibration sensor;

[0046] Step S2: setting the amplification factor, amplifying the collected weak vibration signal through the power amplifier control algorithm to improve the signal-to-noise ratio of the signal, and using the adaptive denoising algorithm to denoise the amplified signal to obtain a denoised signal;

[0047] In the present invention, the signals collected by the vibration sensor are very weak, and the amplitude of these signals is not enough to be directly used for analysis and control. By amplifying these signals through the power amplifier control algorithm, the amplitude of the signal can be increased, making it easier to be recognized and processed by subsequent signal processing and analysis algorithms; secondly, the amplification process can also improve the signal-to-noise ratio of the signal. While amplifying the signal, the background noise and interference will also be amplified, but because the amplitude of the signal itself is increased more, the signal-to-noise ratio will be improved, which will help to further improve the quality of the signal.

[0048] Furthermore, the specific formula of the power amplifier control algorithm is:

[0049]

[0050] Where, Gain represents the gain of the power amplifier, that is, the amplification factor, V out Represents the output voltage of the power amplifier, V in It represents the input voltage of the power amplifier, that is, the collected weak vibration signal.

[0051] In the present invention, the adaptive denoising algorithm is used to remove noise from the amplified signal to improve the quality of the signal. Usually, adaptive adjustment is performed according to the characteristics of the signal and the noise to achieve the best denoising effect.

[0052] Step S3: Perform spectrum analysis and time domain analysis on the denoised signal to extract characteristic information of the signal, combine the results of multi-source data fusion, predict the vibration trend and risk of the building through a deep learning model, formulate a power amplifier control strategy based on the prediction results and the extracted signal characteristic information, apply the formulated power amplifier control strategy to the actual power amplifier control system, and output the processed signal data to a display device for display;

[0053] Step S4: Set the vibration threshold and warning conditions. When the vibration of the building exceeds the vibration threshold or meets the warning conditions, the warning mechanism is automatically triggered to send warning information to the user. At the same time, emergency response measures are automatically initiated according to the warning level and the actual situation of the building.

[0054] Example 2

[0055] See also Figure 3-Figure 4 In this embodiment, the specific steps of step S3 include:

[0056] S3.1: Use Fourier transform to perform spectrum analysis on the denoised signal to obtain the signal spectrum diagram, and extract frequency domain features from the spectrum diagram. At the same time, observe the waveform, peak value, and period of the denoised signal in the time domain to extract time domain features, where frequency domain features include: spectrum distribution and bandwidth;

[0057] S3.2: Introduce multi-dimensional environmental data, fuse frequency domain features and time domain features with multi-dimensional environmental data to obtain fused data, pass the fused data as input to the input layer of the neural network model, and pass the input data to the hidden layer starting from the input layer. At the same time, in each layer, calculate the output of each neuron according to the weight and bias of the current layer. The calculation formula of the neuron output result is:

[0058]

[0059] in, represents the output of the i-th neuron in the a-th layer, A da (·) represents the adaptive activation function, represents the weight from the jth neuron in the (a-1)th layer to the ith neuron in the ath layer, A te (·) represents the function of attention weight, represents the output of the jth neuron in the (a-1)th layer, represents the bias term of the i-th neuron in the a-th layer, Represents the output of the i-th neuron in the (a-1)-th layer.

[0060] Multi-dimensional environmental data includes: temperature, humidity, and wind speed.

[0061] Furthermore, the adaptive activation function A da The expression of (·) is:

[0062]

[0063] in, Represents a trainable parameter whose initial value is set to a small positive number, such as 0.01, so that it can be dynamically adjusted during training. Represents input data, when When , the function output is when When , the function output is The present invention uses an adaptive activation function to solve the problem of activation function The problem of gradient being 0 when the neural network is trained is solved, which avoids the phenomenon of neuron death and introduces trainable parameters. This enables the model to adaptively adjust the shape of the activation function according to different data sets and tasks, thereby improving the flexibility and generalization ability of the model.

[0064] S3.3: Use the activation function to perform nonlinear transformation on the output of the neuron, and pass the fused data after nonlinear transformation to the next layer until it reaches the output layer. At the same time, at the output layer, the vibration trend and risk results of the building predicted by the neural network are obtained.

[0065] Furthermore, the prediction process of the neural network model can be summarized as follows:

[0066] (1) Load model parameters: load the weights and bias items of the pre-trained model from storage;

[0067] (2) Input data: The fused data to be predicted is passed as input to the input layer of the model;

[0068] (3) Layer-by-layer calculation: Starting from the input layer, the output of each neuron is calculated layer by layer according to the weights, bias terms, and activation functions until the output layer is reached;

[0069] (4) Output results: Obtain the final prediction results from the output layer, namely the vibration trend and risk of the building.

[0070] S3.4: Based on the prediction results of the neural network model and the extracted frequency domain characteristics and time domain characteristics information, analyze whether the power amplifier is working in the saturation region and whether the performance index requirements are met, and formulate the power amplifier control strategy through the control algorithm;

[0071] Furthermore, if the power amplifier operates in the saturation region, the output power is close to its maximum capability, but the distortion may increase and the efficiency may decrease, so it is necessary to reduce the input power to reduce the distortion and use pre-distortion technology to compensate for the nonlinear distortion generated by the amplifier in the saturation region;

[0072] If the power amplifier does not operate in the saturation region, such as the linear region, although the distortion may be low, the output power has not reached the maximum and the efficiency is not optimal. In this case, the input power needs to be increased, but care must be taken to avoid entering the saturation region. At the same time, the gain and phase of the amplifier should be adjusted to match the system requirements to ensure the accuracy and integrity of the signal.

[0073] S3.5: converting the formulated power amplifier control strategy into execution instructions and applying it to the actual power amplifier control system to execute the control strategy by adjusting the parameters of the power amplifier;

[0074] Furthermore, the specific steps of converting the formulated power amplifier control strategy into execution instructions include:

[0075] (1) Clearly define the adjustment parameters: Clearly define the power amplifier parameters that need to be adjusted, such as gain, bias current, and bias voltage;

[0076] (2) Calculate the adjustment amount: Calculate the parameter value that needs to be adjusted based on the control strategy and the current state of the system. For example, if the control strategy is error-based feedback control, it is necessary to calculate the error value and calculate the gain adjustment amount based on the error value.

[0077] (3) Converting into execution instructions: Converting the calculated parameter adjustment amount into an execution instruction that can be understood by the power amplifier control system, which involves converting the digital value into a specific control signal or code. For example, if the power amplifier control system receives instructions through serial communication, it is necessary to convert the parameter adjustment amount into a format specified in the serial communication protocol, such as ASCII code or hexadecimal code;

[0078] (4) Sending execution instructions: Sending the converted execution instructions to the power amplifier control system through the serial port, Ethernet, SPI, or I2C. When sending instructions, the reliability and stability of the communication interface must be ensured to avoid data loss or errors.

[0079] (5) Verify command execution: After sending the execution command, verify whether the power amplifier has adjusted the corresponding parameters according to the command. This is mainly achieved by reading the status information of the power amplifier or measuring its output signal. If it is found that the parameter adjustment is incorrect or the system does not respond to the command, it is necessary to recheck the calculation process, command conversion and sending process, and make necessary adjustments;

[0080] (6) Recording and feedback: Record the process and results of parameter adjustment, and feed back the execution results to the control system or operator.

[0081] S3.6: Output the processed spectrum diagram, time domain waveform, and prediction results to a display device, and display the vibration conditions and control effects of the building in an animated form.

[0082] The specific steps in formulating the power amplifier control strategy in S3.4 include:

[0083] S3.41: According to the performance analysis results, the control target of the power amplifier is set. The formula form of the control target is:

[0084]

[0085] Among them, P 0，max Indicates the maximum output power, U0 indicates the effective value of the output voltage, I0 indicates the effective value of the output current, min(THD k ) represents the minimum total harmonic distortion, U k represents the voltage amplitude of the kth harmonic, ζ represents the harmonic frequency, U1 is the fundamental voltage amplitude, η max represents the maximum energy conversion efficiency of the power amplifier, P1 represents the signal power obtained by the load, V cc Indicates the power supply voltage, I avg represents the average current flowing through the power supply, ΔT represents the parameter fluctuation range, T max 、T min They represent the maximum and minimum parameter values ​​during the working process, BW represents bandwidth, and f H 、f L They represent the upper cutoff frequency and the lower cutoff frequency respectively, SNR represents the signal-to-noise ratio, S represents the signal power, and N represents the noise power.

[0086] S3.42: Based on the prediction results and real-time feedback signals of the neural network model, a predictive control algorithm is used to adjust the control parameters of the power amplifier, and the performance of the predictive control algorithm is evaluated according to the adjustment results. The specific formula for evaluating the performance of the predictive control algorithm is:

[0087]

[0088] Where Q represents the performance indicator of the predictive control algorithm, [m, n] represents the time range for evaluating the tracking error, and w q represents the weight of each prediction time step, t represents time, x e (t+q) represents the expected response at time (t+q), q represents the time step, and x mol(t+q) represents the response of the neural network model at time (t+q), β1 and β2 represent the weight factors of the control signal increment and smoothness, respectively, and n u represents the time range for evaluating the control increment, u′(t+q-1), u′(t+q-2), and u′(t+q) represent the control signals obtained by the predictive control algorithm at (t+q-1), (t+q-2), and (t+q), respectively, and λ l Represents the weight factor of the constraint condition, g l (u′) represents the constraint condition of the control signal u′, r represents the number of the constraint conditions, and the calculation method of the control signal u′(·) in the present invention adopts the PID control algorithm, which is the prior art content in this field and is not the inventive solution of the present application, and will not be described in detail here;

[0089] S3.43: Use the gradient descent optimization control algorithm to globally optimize the control parameters and find the optimal parameter combination;

[0090] Furthermore, the specific steps of S3.43 include:

[0091] (1) Set initial parameters: determine the initial parameter combination and step size;

[0092] (2) Calculate the gradient: Calculate the gradient of the current parameter combination based on the objective function;

[0093] (3) Update parameters: Update the parameters in the opposite direction of the gradient. The formula is:

[0094]

[0095] in, They represent the parameters of the (s+1)th and sth iterations respectively, θ represents the step size, Represents the objective function exist The gradient at

[0096] (4) Determine whether the termination condition is met: set a parameter change threshold. If the preset number of iterations is reached or the parameter change is less than the parameter change threshold, stop the iteration; otherwise, return to step (2);

[0097] (5) Output results: Output the optimized parameter combination.

[0098] S3.44: Integrate the prediction model, control algorithm, and optimization algorithm into a unified control system for comprehensive monitoring and evaluation, and use the evaluation results as input to the control algorithm and adjustment strategy to achieve closed-loop feedback control.

[0099] Example 3

[0100] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a weak signal processing method for power amplifier control when executing the computer program.

[0101] A computer-readable storage medium stores computer instructions, which, when executed, execute steps of a weak signal processing method for power amplifier control.

[0102] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.

Claims

1. A weak signal processing method for power amplifier control, characterized in that: include: Step S1: Use a multi-axis vibration sensor to monitor the building in real time, connect the sensor to a data acquisition device, and collect weak signals generated by vibration in real time through wireless transmission; Step S2: setting the amplification factor, amplifying the collected weak vibration signal through the power amplifier control algorithm, and denoising the amplified signal using the adaptive denoising algorithm to obtain a denoised signal; Step S3: Perform spectrum analysis and time domain analysis on the denoised signal to extract characteristic information of the signal, combine the results of multi-source data fusion, predict the vibration trend and risk of the building through a deep learning model, formulate a power amplifier control strategy based on the prediction results and the extracted signal characteristic information, apply the formulated power amplifier control strategy to the actual power amplifier control system, and output the processed signal data to a display device for display; Step S4: Setting a vibration threshold and warning conditions. When the vibration of the building exceeds the vibration threshold or meets the warning conditions, the warning mechanism is automatically triggered to send a warning message to the user. At the same time, emergency response measures are automatically initiated according to the warning level and the actual situation of the building. The specific steps of step S3 include: S3.1: Use Fourier transform to perform spectrum analysis on the denoised signal to obtain the signal spectrum diagram, and extract frequency domain features from the spectrum diagram. At the same time, observe the waveform, peak value, and period of the denoised signal in the time domain to extract time domain features; S3.2: Introduce multi-dimensional environmental data, fuse frequency domain features and time domain features with multi-dimensional environmental data to obtain fused data, pass the fused data as input to the input layer of the neural network model, and pass the input data to the hidden layer starting from the input layer. At the same time, in each layer, calculate the output of each neuron according to the weight and bias term of the current layer; S3.3: Use the activation function to perform nonlinear transformation on the output of the neuron, and pass the fused data after nonlinear transformation to the next layer until it reaches the output layer. At the same time, at the output layer, the vibration trend and risk results of the building predicted by the neural network are obtained; The specific steps of step S3 also include: S3.4: Based on the prediction results of the neural network model and the extracted frequency domain characteristics and time domain characteristics information, analyze whether the power amplifier is working in the saturation region and whether the performance index requirements are met, and formulate the power amplifier control strategy through the control algorithm; S3.5: converting the formulated power amplifier control strategy into execution instructions and applying it to the actual power amplifier control system to execute the control strategy by adjusting the parameters of the power amplifier; S3.6: Output the processed spectrum diagram, time domain waveform, and prediction results to a display device, and display the vibration conditions and control effects of the building in an animated form.

2. A weak signal processing method for power amplifier control as claimed in claim 1, characterized in that: The calculation formula of the neuron output result in S3.2 is: in, represents the output of the i-th neuron in the a-th layer, A da (·) represents the adaptive activation function, represents the weight from the jth neuron in the (a-1)th layer to the ith neuron in the ath layer, A te (·) represents the function of attention weight, represents the output of the jth neuron in the (a-1)th layer, represents the bias term of the i-th neuron in the a-th layer, Represents the output of the i-th neuron in the (a-1)-th layer.

3. A weak signal processing method for power amplifier control as claimed in claim 2, characterized in that: The specific steps of formulating the power amplifier control strategy in S3.4 include: S3.41: Setting a control target of the power amplifier according to the performance analysis result; S3.42: Based on the prediction results of the neural network model and the real-time feedback signal, use the predictive control algorithm to adjust the control parameters of the power amplifier, and evaluate the performance of the predictive control algorithm based on the adjustment results; S3.43: Use the gradient descent optimization control algorithm to globally optimize the control parameters and find the optimal parameter combination; S3.44: Integrate the prediction model, control algorithm, and optimization algorithm into a unified control system for comprehensive monitoring and evaluation, and use the evaluation results as input to the control algorithm and adjustment strategy to achieve closed-loop feedback control.

4. A weak signal processing method for power amplifier control as claimed in claim 3, characterized in that: The control objective in S3.41 is expressed as: Among them, P 0,max Indicates the maximum output power, U0 indicates the effective value of the output voltage, I0 indicates the effective value of the output current, min(THD k ) represents the minimum total harmonic distortion, U k represents the voltage amplitude of the kth harmonic, ζ represents the harmonic frequency, U1 is the fundamental voltage amplitude, η max represents the maximum energy conversion efficiency of the power amplifier, P1 represents the signal power obtained by the load, V cc Indicates the power supply voltage, I avg represents the average current flowing through the power supply, ΔT represents the parameter fluctuation range, T max , T min They represent the maximum and minimum parameter values ​​during the working process, BW represents bandwidth, and f H 、f L They represent the upper cutoff frequency and the lower cutoff frequency respectively, SNR represents the signal-to-noise ratio, S represents the signal power, and N represents the noise power.

5. A weak signal processing method for power amplifier control as claimed in claim 4, characterized in that: The specific formula for evaluating the performance of the predictive control algorithm in S3.42 is: Where Q represents the performance indicator of the predictive control algorithm, [m,n] represents the time range for evaluating the tracking error, and w q represents the weight of each prediction time step, t represents time, x e (t+q) represents the expected response at time (t+q), q represents the time step, and x mol (t+q) represents the response of the neural network model at time (t+q), β1 and β2 represent the weight factors of the control signal increment and smoothness respectively, and n u represents the time range for evaluating the control increment, u′(t+q-1), u′(t+q-2), and u′(t+q) represent the control signals obtained by the predictive control algorithm at (t+q-1), (t+q-2), and (t+q), respectively, and λ l Represents the weight factor of the constraint condition, g l (u′) represents the constraint condition of the control signal u′, and r represents the number of constraint conditions.

6. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a weak signal processing method for power amplifier control described in any one of claims 1 to 5 are implemented.

7. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, the steps of a weak signal processing method for power amplifier control as described in any one of claims 1 to 5 are executed.

Images