Physical experiment data acquisition method and system based on multi-modal sensor

Abstract

The invention discloses a physical experiment data acquisition method and system based on a multi-modal sensor, and relates to the technical field of data fusion. The method comprises the following steps: acquiring main frequency band information and interference characteristic information, outputting an anti-interference instruction set in combination with a preset instruction mapping rule, acquiring a mixed data set for verification and a sensing data set for verification so as to obtain effect evaluation indexes of various instructions in the anti-interference instruction set, and comparing the effect evaluation indexes with preset effect thresholds of the effect evaluation indexes. If at least one effect evaluation index which does not reach the preset effect threshold value exists, adjusting the corresponding instruction until each effect evaluation index reaches the preset effect threshold value, and obtaining the adjusted frequencies of different instructions to update the preset step length when the instruction adjusts the behavior; environmental electromagnetic interference can be effectively suppressed, physical experiment data quality and measurement repeatability are improved, and experiment cost is reduced.

Description

Technical Field

[0001] This application relates to the field of data fusion technology, specifically to a method and system for acquiring physical experimental data based on multimodal sensors. Background Technology

[0002] In physical experiments involving weak signal detection, the experimental environment is often subject to electromagnetic interference from wireless communication devices, switching power supplies, and mobile terminals. This interference is characterized by high frequency, intermittency, and directionality, which causes the target physical signal to be contaminated during the acquisition process. Therefore, under normal experimental conditions, the quality of physical experimental data fluctuates significantly, the repeatability of measurement results is poor, and it is difficult to meet the requirements of high-precision experiments.

[0003] Furthermore, existing technologies lack real-time sensing and dynamic response mechanisms for environmental electromagnetic noise, and cannot adjust acquisition strategies according to changes in interference during data acquisition. This results in passive and inefficient data acquisition and low resource utilization. These problems make the acquisition of high-quality physical experimental data heavily dependent on special shielded environments, leading to excessively high experimental thresholds and hindering the widespread application of experiments. To address these issues, existing technologies urgently need improvement. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for acquiring physical experimental data based on multimodal sensors, which can effectively suppress environmental electromagnetic interference, improve the quality and repeatability of physical experimental data, and reduce experimental costs.

[0005] The objective of this application can be achieved through the following technical solution: Firstly, a method for acquiring physical experimental data based on multimodal sensors, comprising the following steps:

[0006] Simultaneously acquire the original hybrid dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise.

[0007] Spectral analysis is performed on the original mixed dataset to obtain the main frequency band information of the target physical signal, and time-frequency analysis is performed on the auxiliary sensing dataset to obtain interference characteristic information reflecting environmental electromagnetic interference.

[0008] Based on the main frequency band information and the interference feature information, combined with the preset instruction mapping rules, an anti-interference instruction set containing multiple instructions is output;

[0009] The original hybrid dataset and auxiliary sensing dataset within a preset acquisition period after the execution of the anti-interference instruction set are obtained, and these are used as the verification hybrid dataset and verification sensing dataset, respectively. The effectiveness evaluation index of each instruction in the anti-interference instruction set is obtained by combining the main frequency band information and the interference feature information.

[0010] Each effect evaluation index is compared with its preset effect threshold. If each effect evaluation index reaches its preset effect threshold, the anti-interference instruction set remains unchanged.

[0011] If at least one effect evaluation index fails to reach its preset effect threshold, adjust the corresponding instruction in the anti-interference instruction set until all effect evaluation indices reach their preset effect thresholds.

[0012] Obtain the adjusted frequency of different instructions within the preset optimization period, and update the preset step size when adjusting the instruction behavior according to the adjusted frequency of different instructions.

[0013] Secondly, the physical experiment data acquisition module based on multimodal sensors includes the following modules:

[0014] The data acquisition module is used to synchronously acquire the original mixed dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise.

[0015] The data processing module is used to perform spectral analysis on the original mixed dataset to obtain the main frequency band information of the target physical signal, and to perform time-frequency analysis on the auxiliary sensing dataset to obtain interference characteristic information reflecting environmental electromagnetic interference.

[0016] The instruction mapping module is used to output an anti-interference instruction set containing multiple instructions based on the main frequency band information and the interference feature information combined with preset instruction mapping rules.

[0017] The effect evaluation module is used to acquire the original mixed dataset and auxiliary sensing dataset within a preset acquisition period after the execution of the anti-interference instruction set, and use them as the verification mixed dataset and verification sensing dataset respectively, and combine the main frequency band information and the interference feature information to obtain the effect evaluation index of each instruction in the anti-interference instruction set;

[0018] The instruction adjustment module is used to compare each effect evaluation index with its preset effect threshold. If all effect evaluation indices reach their preset effect thresholds, the anti-interference instruction set remains unchanged. If at least one effect evaluation index fails to reach its preset effect threshold, the corresponding instruction in the anti-interference instruction set is adjusted until all effect evaluation indices reach their preset effect thresholds.

[0019] The feedback optimization module is used to obtain the adjustment frequency of different instructions within a preset optimization period, and update the preset step size when adjusting the instruction behavior according to the adjustment frequency of different instructions.

[0020] Thirdly, a computer storage medium stores computer-executable instructions, which, when executed, implement the physical experiment data acquisition method based on multimodal sensors described in the first aspect.

[0021] Compared with the prior art, the beneficial effects of this application are:

[0022] This application dynamically acquires the original mixed dataset and the auxiliary sensing dataset, analyzes the main frequency band information and interference characteristics, and outputs and optimizes the anti-interference instruction set. Combined with the effect evaluation and adaptive adjustment mechanism, it effectively solves the problems of data pollution and poor measurement repeatability caused by electromagnetic interference. It can effectively suppress environmental electromagnetic interference, improve the quality of physical experimental data and measurement repeatability, and reduce experimental costs. Attached Figure Description

[0023] Figure 1 This is a schematic diagram illustrating the steps of the physical experiment data acquisition method based on multimodal sensors according to this application;

[0024] Figure 2 This is a schematic diagram of the physical experiment data acquisition module based on multimodal sensors according to this application. Detailed Implementation

[0025] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components described and shown in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but only to illustrate selected embodiments of this application.

[0026] Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this application. It should be noted that similar reference numerals and letters in the following figures indicate similar items. Therefore, once an item has been defined in one figure, it does not need to be further defined and explained in subsequent figures. The terms "first", "second", etc. are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0027] During the acquisition of physical experimental data, especially in scenarios involving weak signal detection, noise signals introduced by high-frequency, intermittent electromagnetic interference sources in the laboratory environment (such as wireless LAN devices, switching power supplies, and mobile communication terminals) mix with the target physical signal, resulting in interference components in the raw data. Existing static electromagnetic shielding measures or fixed-parameter filters cannot adapt to the dynamic changes in interference characteristics, leading to unstable signal-to-noise ratios of the target physical signal and affecting the reliability of the data and the accuracy of the experimental results. This problem is particularly prominent in ordinary laboratory environments without professional shielding, making it difficult to consistently achieve high-quality data acquisition.

[0028] For example, in a Hall voltage measurement experiment, when a multi-channel data acquisition module was used to record weak Hall effect signals, the wireless router in the laboratory periodically emitted a 2.4 GHz frequency band signal, while the switching power supply generated 50 Hz harmonic interference. Within the preset acquisition period, the original mixed dataset was observed to have the target physical signal spectrum covered by the interference frequency band, and the auxiliary sensing dataset recorded the temporal regularity and directional changes of the interference pulses. Due to the intermittent and dynamic characteristics of the interference, the signal and noise components could not be effectively separated by the fixed filter, resulting in transient noise spikes and baseline drift in the acquired voltage data. The consistency of data from multiple experiments was reduced, and the experimental process required repeated parameter adjustments.

[0029] If the aforementioned problems are not addressed, the continued impact of electromagnetic interference will further deteriorate the signal-to-noise ratio of physical experimental data, making it difficult to accurately identify and extract weak signal features. This not only increases the risk of experimental failure but may also introduce systematic errors, affecting the scientific validity and reproducibility of experimental results. In the long run, this problem limits the possibility of conducting high-precision physical experiments under ordinary laboratory conditions, forcing researchers to rely on expensive shielding facilities, thereby increasing experimental costs and limiting the widespread availability of experiments.

[0030] Therefore, such as Figure 1 As shown, this application provides a method for acquiring physical experimental data based on multimodal sensors, including the following steps:

[0031] Simultaneously acquire the original hybrid dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise.

[0032] Spectral analysis is performed on the original mixed dataset to obtain the main frequency band information of the target physical signal, and time-frequency analysis is performed on the auxiliary sensing dataset to obtain interference characteristic information reflecting environmental electromagnetic interference.

[0033] Based on the main frequency band information and the interference feature information, combined with the preset instruction mapping rules, an anti-interference instruction set containing multiple instructions is output;

[0034] The original hybrid dataset and auxiliary sensing dataset within a preset acquisition period after the execution of the anti-interference instruction set are obtained, and these are used as the verification hybrid dataset and verification sensing dataset, respectively. The effectiveness evaluation index of each instruction in the anti-interference instruction set is obtained by combining the main frequency band information and the interference feature information.

[0035] Each effect evaluation index is compared with its preset effect threshold. If each effect evaluation index reaches its preset effect threshold, the anti-interference instruction set remains unchanged.

[0036] If at least one effect evaluation index fails to reach its preset effect threshold, adjust the corresponding instruction in the anti-interference instruction set until all effect evaluation indices reach their preset effect thresholds.

[0037] Obtain the adjusted frequency of different instructions within the preset optimization period, and update the preset step size when adjusting the instruction behavior according to the adjusted frequency of different instructions.

[0038] The above technical solution achieves comprehensive perception of interference sources by simultaneously acquiring target physical signals and environmental noise through multi-modal sensors. Compared with traditional methods that rely on only a single sensor or fixed-parameter filters, this solution can more accurately identify the main frequency band of the target physical signal and the complex characteristics of environmental electromagnetic interference. This solution transforms the perceived main frequency band information and interference characteristics into a dynamic anti-interference instruction set through instruction mapping rules, including various instructions such as filtering parameters, acquisition windows, and shielding positions. This contrasts with the limitations of static shielding or fixed-parameter filters in existing technologies, which cannot adapt to dynamically changing interference environments. This solution can generate customized anti-interference strategies based on real-time environmental changes.

[0039] Furthermore, this scheme introduces a closed-loop feedback mechanism for effect evaluation and command adjustment, which quantitatively evaluates the execution effect of commands and dynamically adjusts command parameters. This enables the system to continuously optimize anti-interference strategies until a preset effect threshold is reached. Compared with the shortcomings of existing technologies that lack feedback mechanisms and cannot self-optimize, this scheme can ensure a stable improvement in the signal-to-noise ratio of the target physical signal in an unshielded ordinary laboratory environment. Finally, by introducing a mechanism to update the preset step size based on the frequency at which commands are adjusted, this scheme achieves adaptive optimization of the adjustment behavior. This allows the system to dynamically adjust its adjustment sensitivity according to the activity level of commands in actual applications, thereby ensuring adjustment efficiency while avoiding over-adjustment or under-adjustment. This intelligent optimization process transforms data acquisition from passive recording to an active optimization process, significantly reducing the stringent environmental requirements for high-quality experiments.

[0040] It should be further explained that, in the specific implementation process, the process of simultaneously acquiring the original hybrid dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise, includes:

[0041] The original mixed dataset refers to a time-series data set collected at continuous time points, reflecting the superposition state of the target physical signal and background electromagnetic interference. The data at each time point is typically represented as a voltage value or a digitized numerical value. The entire sequence is arranged according to a fixed sampling interval, completely recording the dynamic changes of the target physical signal and the interference it experiences within a complete experimental action or observation period. The core characteristic of this dataset is that the signal and interference are mixed in both the time and frequency domains, making them impossible to separate through simple observation.

[0042] This data acquisition is achieved through a high-precision, high-sensitivity main sensor. In the experimental setup, the main sensor's detection unit (such as a probe, coil, or photoelectric element) is placed within the field or path of the physical quantity to be measured, ensuring a stable connection to the experimental apparatus. When the experiment begins with a preset action or change, the main sensor is activated for continuous data acquisition. For example, a 24-bit high-resolution analog-to-digital converter (ADC) data acquisition card can be used: its input is connected to the target physical signal line output by the preamplifier, continuously sampling and digitizing the voltage signal at a preset high sampling rate (e.g., 1 kS / s to 1 MS / s, determined based on the signal frequency). The data output is a continuous one-dimensional voltage sequence file with timestamps. This device has a large dynamic range and low noise floor, making it suitable for capturing weak signals at the microvolt level.

[0043] The auxiliary sensing dataset refers to a multimodal data set synchronously collected by a group of auxiliary sensors, specifically designed to characterize the electromagnetic noise characteristics (such as intensity, spectrum, direction, and time-varying patterns) of the experimental environment. Its core function is to provide independent spatiotemporal and spectral feature information about the interference source, providing a basis for subsequent identification and suppression of interference from mixed data.

[0044] The acquisition is achieved through an auxiliary sensor array that is strictly synchronized with the main sensor. This array typically includes the following sensor units: 1) Broadband electromagnetic field probes: Multiple probes are arranged in different directions (such as east, west, and north) around the experimental area according to the principle of spatial interferometry positioning, at a certain distance from the main sensor (such as 0.5 meters to 1 meter), to capture the changes in broadband (such as from 100kHz to 2.4GHz) electromagnetic field intensity in a specific direction in space. 2) Current probes: Clamped at the power supply line entrance to the experimental device in the laboratory, used to couple and acquire common-mode and differential-mode noise conducted through the power line.

[0045] The main sensor and all auxiliary sensors are simultaneously activated to begin data acquisition via a common external hardware trigger signal or software commands based on a high-precision network clock protocol. Wideband probes typically have a built-in high-speed ADC, capable of directly outputting digitized spectral data or high-speed time-domain waveforms; current probe outputs are connected to another synchronous acquisition channel. All auxiliary sensor data are ultimately strictly time-aligned with the main sensor data and can be calibrated to a common physical dimension reference.

[0046] In other embodiments, this application further proposes to perform a Fourier transform on the original mixed dataset to obtain its power spectral density, and to take the continuous frequency range in the power spectral density that is greater than a first preset frequency threshold as the main frequency band information of the target physical signal; to perform a short-time Fourier transform on the auxiliary sensing dataset to obtain time-spectrum data, and to extract interference feature information including interference frequency band information, interference pulse information, and interference direction information based on the time-spectrum data.

[0047] From the frequency dimension of the time-spectrum data, the frequency range with average energy exceeding the second preset frequency threshold is determined as the interference frequency band information; from the time dimension of the time-spectrum data, its envelope signal is extracted, and by performing peak detection and statistical analysis on the envelope signal, interference pulse information characterizing the time pattern of interference pulses is obtained; cross-correlation analysis or energy ratio calculation is performed on the time-spectrum data of multiple auxiliary sensing datasets from different spatial orientations to obtain the time delay difference or intensity difference of the same interference frequency band signal arriving at different probes, and the direction of arrival information of interference is obtained based on the time delay difference or intensity difference through the direction of arrival estimation algorithm.

[0048] Specifically, the Fourier transform is a mathematical method that converts a time-domain signal into a frequency-domain signal, revealing the frequency components contained in the signal. In practical applications, the Fast Fourier Transform (FFT) algorithm is typically used for efficient computation. The Fourier transform yields the power spectral density of the signal, and this density curve visually reflects the distribution of signal energy at different frequencies. The Short-Time Fourier Transform (SFT) is a method for analyzing the time-frequency characteristics of time-varying signals. It performs a Fourier transform after windowing and frame-by-frame processing of the signal to obtain time-spectrum data. This time-spectrum data simultaneously displays the changes in the signal's frequency components over time, making it particularly effective for analyzing non-stationary electromagnetic interference signals.

[0049] Based on time-spectrum data, various interference characteristics can be extracted. Interference frequency band information refers to the frequency range occupied by the electromagnetic interference signal, which can be obtained through frequency range energy analysis or pattern recognition of the time-spectrum data. Interference pulse information refers to instantaneous, high-energy electromagnetic interference events, which can be identified by detecting instantaneous energy peaks in the time-spectrum data or setting energy thresholds. Interference direction information indicates the spatial location or direction of the electromagnetic interference source, which typically requires combining multi-mode sensor array data and estimation using algorithms such as beamforming, time difference of arrival (TDOA), or angle of arrival (DOA).

[0050] The above technical solution clarifies how to obtain the power spectral density by performing a Fourier transform on the original mixed dataset, and how to accurately identify the main frequency band information of the target physical signal based on the power spectral density, thus avoiding the ambiguity of main frequency band information identification. Simultaneously, by performing a short-time Fourier transform on the auxiliary sensing dataset to obtain time-spectrum data, and based on this, extracting interference frequency band information, interference pulse information, and interference direction-of-arrival information, a more comprehensive and in-depth understanding of the characteristics of environmental electromagnetic interference is achieved. This precise and multi-dimensional information extraction provides a solid foundation for the subsequent generation of anti-interference commands, enabling the anti-interference command set to more accurately suppress interference sources, thereby significantly improving the anti-interference capability and data quality of physical experimental data acquisition and ensuring the reliability of experimental data.

[0051] In other embodiments, this application further proposes a pre-defined mapping function that uses main frequency band information and interference characteristic information as input data and different commands as output data as the command mapping rule. The main frequency band information and the interference frequency band information are input into the command mapping rule to output filter parameter commands, and the interference pulse information and the interference direction information are input into the command mapping rule to output acquisition window commands and shielding position commands respectively. The anti-interference command set is a collection of different commands output based on the command mapping rule.

[0052] The process involves inputting the main frequency band information and interference frequency band information into the instruction mapping rule to output filter parameter instructions. This process aims to intelligently generate parameters for digital or analog filters based on the frequency range of the target physical signal and the frequency range of the electromagnetic interference signal. For example, when an overlap between the interference frequency band and the main frequency band is detected, the mapping rule can output the center frequency and bandwidth parameters of a narrowband notch filter; when the interference frequency band is far from the main frequency band, it may output the cutoff frequency parameter of a bandpass filter.

[0053] The process of inputting interference pulse information and interference source information into the command mapping rule to output acquisition window instructions and shielding position instructions, respectively, leverages the transient and spatial characteristics of interference to optimize data acquisition and physical shielding. For example, when a high-energy interference pulse is detected, the command mapping rule can output an instruction to the data acquisition module to pause or adjust the sampling rate when the pulse appears, thereby generating a corresponding acquisition window. When an interference source in a specific direction is detected, the command mapping rule can output an instruction to the movable physical shielding device to move to a specific position and adopt a specific azimuth angle to block or attenuate interference from that direction, thereby generating a shielding position instruction.

[0054] The anti-interference instruction set is a collection of different instructions output based on instruction mapping rules. This means that the anti-interference strategy generated by the system is not a single instruction, but a set of instructions working in tandem. This set of instructions may include filtering parameter instructions, acquisition window instructions, shielding position instructions, etc., which together constitute a comprehensive anti-interference scheme. This aggregated instruction output method enables the system to take multi-dimensional and comprehensive countermeasures against electromagnetic interference of different types and sources, thereby more effectively protecting the acquisition quality of the target physical signal.

[0055] Through the above technical solution, this application can transform environmental electromagnetic interference sensing information into a specific, executable set of anti-interference instructions through preset instruction mapping rules. This transformation mechanism enables the system to intelligently generate multi-dimensional anti-interference strategies, such as filtering parameter instructions, acquisition window instructions, and shielding position instructions, based on the main frequency band information of the target physical signal and various characteristics of environmental electromagnetic interference (including frequency band, pulse, and direction of arrival). This avoids the inefficiency and uncertainty of manual judgment and adjustment, significantly improving the automation and intelligence level of the physical experiment data acquisition process. By closely integrating sensing and control, the system can take customized countermeasures against interference of different types and sources, thereby effectively suppressing the contamination of the target physical signal by environmental electromagnetic noise, greatly improving the signal-to-noise ratio and reliability of the acquired data, and ensuring the accuracy of the physical experiment results.

[0056] In some other embodiments, this application further proposes that after controlling the physical shielding device to move to the corresponding position according to the shielding position command, the original mixed dataset and the auxiliary sensing dataset under the corresponding sampling window within the preset acquisition period are obtained according to the acquisition window command, and then the obtained original mixed dataset and auxiliary sensing dataset are filtered according to the filtering parameters in the filtering parameter command to obtain the verification mixed dataset and the verification sensing dataset.

[0057] The shielding position command is an instruction to move a physical shielding device to a specific location. The concept is to optimize the suppression of electromagnetic interference from a specific direction or source by changing the spatial position of the shield. A physical shielding device is a device designed to reduce the impact of environmental electromagnetic interference on sensors by physically blocking or absorbing electromagnetic waves. Possible implementations include: a Faraday cage or shield made of conductive materials (such as copper or aluminum), whose position can be adjusted electrically or manually; or a structure containing absorbing materials (such as ferrite or carbon-based materials) to absorb electromagnetic waves in a specific frequency range, which can be moved by a robotic arm or linear actuator.

[0058] The acquisition window instruction is a command that instructs the data acquisition module to select a specific acquisition window for data sampling within a preset acquisition period. The concept is to improve the quality of effective data acquisition by avoiding interference pulses or high-noise periods. This is achieved through a pair of timestamps (including start and end times), defining the specific time period for data sampling within the preset acquisition period. After executing the acquisition window instruction, the data acquisition module only acquires data within the specified sampling window, thereby obtaining the original mixed dataset and the auxiliary sensing dataset that have been time-filtered.

[0059] The filtering parameter command is a set of parameters that instructs the data processing module to use when filtering the acquired data. The concept is to select an appropriate filter type and parameters based on the interference frequency band information to suppress noise within a specific frequency range. This is achieved through a set of digital filter parameters, including center frequency, filter order, and filter type. The data processing module performs digital or analog filtering operations on the original mixed dataset and auxiliary sensing dataset acquired within the sampling window according to the filtering parameters in the filtering parameter command. After physical shielding, sampling window selection, and filtering, the final dataset is used as the verification mixed dataset and verification sensing dataset.

[0060] Through the above technical solution, this application can effectively transform the abstract anti-interference instruction set generated by the instruction mapping module, including shielding position instructions, acquisition window instructions, and filtering parameter instructions, into actual physical operations and data processing steps. This ensures that the verification hybrid dataset and verification sensing dataset used for effect evaluation are obtained under conditions where anti-interference measures are actually applied, thus providing a reliable and representative data foundation for subsequent accurate evaluation of the effects of each instruction. Through the synergistic effect of physical shielding, time window selection, and frequency domain filtering, the impact of environmental electromagnetic interference on the acquisition of physical experimental data can be significantly reduced, improving the signal-to-noise ratio and accuracy of the acquired data, thereby enhancing the effectiveness and robustness of the entire anti-interference data acquisition method.

[0061] In other embodiments, this application proposes using the frequency range overlapping between the main frequency band information and the interference frequency band information as the evaluation frequency band to obtain the original signal-to-noise ratio of the original mixed dataset and the verification mixed dataset within the evaluation frequency band. And verification signal-to-noise ratio Obtain the first effect evaluation index of the filter parameter command. ; Obtain the interference frequency band information and interference pulse information of the verification sensing dataset and use them as the verification frequency band information and verification pulse information respectively, and use the frequency range corresponding to the verification frequency band information as the verification frequency band.

[0062] The number of pulses in the verification sensor dataset within its sampling window is obtained based on the verification pulse information. The number of pulses in the auxiliary sensing dataset within its preset acquisition period is obtained based on the interference pulse information. The second effect evaluation index of the acquisition window command is obtained. , This corresponds to the time length of the sampling window. To correspond to the preset acquisition period time length; obtain the signal power of the auxiliary sensing dataset within the verification frequency band. And obtain the signal power of the sensing dataset used for verification within the same verification frequency band. The third effect evaluation index of the shielding position command is obtained. ;

[0063] The original signal-to-noise ratio The verification uses signal-to-noise ratio ,in, These represent the signal power and noise power of the original mixed dataset within the evaluation frequency band, respectively. These represent the signal power and noise power of the hybrid dataset used for verification within the evaluation frequency band, respectively.

[0064] The evaluation frequency band refers to the frequency range where the main frequency band information of the target physical signal overlaps with the interference frequency band information of the environmental electromagnetic interference. The determination of this frequency band aims to focus on the region where the interaction between signal and interference is most significant, thereby enabling subsequent effect evaluations to more accurately reflect the actual effectiveness of anti-interference measures. The raw signal-to-noise ratio (SNR) represents the signal quality of the raw mixed dataset within the evaluation frequency band before the anti-interference command is executed, i.e., the ratio of signal power to noise power. The verification SNR represents the signal quality of the verification mixed dataset within the same evaluation frequency band after the anti-interference command is executed. The acquisition of these two SNRs aims to provide a quantitative indicator for comparing the changes in signal quality before and after anti-interference processing. The SNR is calculated by performing power spectral density calculations on the raw and verification mixed datasets within the evaluation frequency band, and then estimating the signal power and noise power according to a preset signal and noise separation method.

[0065] The first performance evaluation index is used to quantify the anti-interference effect of the filtering parameter commands. It represents the relative improvement in signal-to-noise ratio. This index can intuitively reflect the degree to which the filtering process improves the signal-to-noise ratio of the target physical signal. For example, if it is positive and large, it indicates that the filtering effect is significant; if it is close to zero or negative, it may mean that the filtering effect is poor or that new problems have been introduced.

[0066] The verification frequency band information and verification pulse information are extracted from the verification sensing dataset after executing anti-jamming commands, and are used to reflect the residual interference frequency band characteristics and pulse interference characteristics after processing, respectively. This information aims to provide a processed interference baseline for comparison with the interference characteristics before processing. (Pulse count) This indicates the number of pulses within the sampling window of the sensor dataset used for verification after executing the acquisition window command. (Pulse count) The first indicator represents the number of pulses in the auxiliary sensing dataset within its preset acquisition period before the acquisition window command is executed. Obtaining these two pulse counts aims to quantify the suppression effect of the acquisition window command on pulse interference. The second effect evaluation index is used to quantify the anti-interference effect of the acquisition window command. It reflects the degree to which the pulse interference density is reduced per unit time by adjusting the sampling window. If it is close to 1, it indicates that the acquisition window command effectively avoids most pulse interference; if it is close to 0 or negative, it may mean that the acquisition window command is ineffective.

[0067] signal power This represents the signal power of the auxiliary sensing dataset within the verification frequency band before the shielding position command is executed. It primarily reflects the intensity of electromagnetic interference from the original environment within that frequency band. (Signal power) The first indicator represents the signal power within the same verification frequency band of the sensing dataset used for verification after executing the shielding position command, primarily reflecting the residual interference intensity after shielding. Obtaining these two signal power values ​​aims to quantify the suppression effect of shielding measures on interference in a specific frequency band. The third effect evaluation index is used to quantify the anti-interference effect of the shielding position command. It represents the relative reduction in electromagnetic interference signal power. This index can intuitively reflect the degree of suppression of environmental electromagnetic interference by the physical shielding device. For example, if it is positive and large, it indicates a significant shielding effect; if it is close to zero or negative, it may mean poor shielding effect or improper shielding position.

[0068] The raw signal-to-noise ratio represents the signal quality of the raw mixed dataset within the evaluation frequency band before any anti-interference measures are applied. Signal power Noise power can be obtained by integrating the spectral components of the original mixed dataset within the evaluation band, or by accumulating the energy of the signal components within the evaluation band. The signal-to-noise ratio (SNR) can be obtained by statistically analyzing the spectral components of the original mixed dataset outside the evaluation band and extrapolating them to the evaluation band, or by identifying and removing signal components within the evaluation band and then calculating the power of the remaining components. The verification SNR represents the signal quality and signal power of the verification mixed dataset within the evaluation band after applying anti-interference measures (especially filtering). and noise power The method of obtaining it is the same as above.

[0069] In some other embodiments, this application further proposes to set corresponding first preset effect threshold, second preset effect threshold, and third preset effect threshold for the first effect evaluation index, the second effect evaluation index, and the third effect evaluation index, respectively. The effect evaluation index includes the first effect evaluation index, the second effect evaluation index, and the third effect evaluation index, and the preset effect threshold includes the first preset effect threshold, the second preset effect threshold, and the third preset effect threshold.

[0070] For the effect evaluation index that does not reach its preset effect threshold, according to the order of the masking position instruction, sampling window instruction, and filtering parameter instruction, a preset step size is updated based on the parameters in the corresponding instruction. The preset step size includes a preset masking step size, a preset sampling step size, and a preset filtering step size. Each update for any instruction is regarded as an instruction adjustment behavior, and the verification hybrid dataset and the verification perception dataset and the effect evaluation index of each instruction are obtained after a single instruction adjustment behavior, until the effect evaluation index of each instruction reaches its corresponding preset effect threshold.

[0071] This clarifies the triggering condition (threshold not reached), the adjustment order, and the adjustment method (parameter updates based on a preset step size) for instruction adjustment. This sequential adjustment mechanism helps to systematically handle multiple non-compliant instructions, avoiding the complex interactions and uncertainties that may result from adjusting multiple parameters simultaneously. The preset step size provides a quantifiable and controllable adjustment increment. The adjustment order is fixed as masking position instruction, sampling window instruction, and filter parameter instruction. When the effect evaluation index of an instruction fails to reach the threshold, the system will first attempt to adjust the masking position instruction; if that still fails, it will then adjust the sampling window instruction, and finally adjust the filter parameter instruction.

[0072] The preset step size includes a preset shielding step size, a preset sampling step size, and a preset filtering step size. This ensures the targeted nature of the command adjustment, with different types of command parameters (such as azimuth, start time, and center frequency) using adjustment units and amplitudes consistent with their physical meanings. The preset shielding step size is an angle value used to adjust the azimuth of the physical shielding device. The preset sampling step size is a time unit used to adjust the start time of the sampling window. The preset filtering step size is a dimensionless numerical value used to adjust the center frequency of the filter.

[0073] The adjustment process is as follows: For the azimuth angle in the shielding position command, simultaneously increase one preset shielding step size and decrease one preset shielding step size, and obtain the adjusted third effect evaluation index respectively. Select the adjustment direction that makes the third effect evaluation index larger for adjustment. For the start time in the sampling window command, simultaneously advance one preset sampling step size and delay one preset sampling step size, and obtain the adjusted second effect evaluation index respectively. Select the adjustment direction that makes the second effect evaluation index larger for adjustment. For the center frequency in the filtering parameter command, simultaneously increase one preset filtering step size and decrease one preset filtering step size, and obtain the adjusted first effect evaluation index respectively. Select the adjustment direction that makes the first effect evaluation index larger for adjustment.

[0074] Each update to any instruction is considered a single instruction adjustment action. The system obtains the validation hybrid dataset and validation perception dataset, along with the effectiveness evaluation index for each instruction, after each adjustment action, until all effectiveness evaluation indices for each instruction reach their corresponding preset effectiveness thresholds. After each instruction adjustment action, the system re-triggers the data acquisition module to obtain the validation hybrid dataset and validation perception dataset, and the effectiveness evaluation module calculates the effectiveness evaluation indices. The adjustment process can be configured with a maximum number of iterations or a maximum adjustment time to prevent infinite loops caused by failure to converge in certain extreme cases.

[0075] In other embodiments, this application further proposes the number of instruction adjustment behaviors based on all instructions within a preset optimization period. And the number of times a single instruction modifies behavior. To obtain the adjusted frequency of this single instruction And obtain the preset step size after the single instruction update. , This is the preset step size before the update of this single instruction. and These are the preset adjustment factor and reference frequency, respectively, both ranging from (0, 1).

[0076] The preset optimization period refers to a pre-defined time period or event interval used for statistical analysis of command adjustment behavior. This period is a fixed time length, such as hourly or daily, and its purpose is to provide a statistical window for calculating the command adjustment frequency. Its time span is much longer than the preset acquisition period. The number of command adjustment behaviors for all commands represents the total number of times all anti-interference commands (including filter parameter commands, acquisition window commands, and shielding position commands) are adjusted within the preset optimization period. This value reflects the activity level and adjustment needs of the entire anti-interference system in the current environment. The number of command adjustment behaviors for a single command represents the number of times a specific anti-interference command is adjusted within the preset optimization period. This value is used to measure the adjustment frequency and importance of that single command within the current optimization period. The adjustment frequency quantifies the adjustment activity of a specific command relative to all commands; a high frequency indicates that the command may require more refined or faster adjustments.

[0077] Through the above technical solution, this application can dynamically adjust the preset step size according to the adjustment frequency of different instructions in actual operation. For instructions that require frequent adjustment, the preset step size can be adaptively increased, thereby accelerating the system convergence to the optimal anti-interference state and significantly improving the adjustment efficiency. For instructions with lower adjustment frequencies, the preset step size can be adaptively decreased, effectively avoiding overshoot and oscillation, and improving the accuracy and stability of adjustment. This adaptive preset step size adjustment mechanism enables physical experimental data acquisition methods to perform anti-interference processing more flexibly, efficiently, and accurately when facing complex and ever-changing environmental electromagnetic interference, thereby ensuring the quality and reliability of data acquisition and shortening the time required to achieve ideal data quality.

[0078] In another implementation, such as Figure 2 As shown, this application also provides a physical experiment data acquisition module based on a multimodal sensor, including the following modules:

[0079] The data acquisition module is used to synchronously acquire the original mixed dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise.

[0080] The data processing module is used to perform spectral analysis on the original mixed dataset to obtain the main frequency band information of the target physical signal, and to perform time-frequency analysis on the auxiliary sensing dataset to obtain interference characteristic information reflecting environmental electromagnetic interference.

[0081] The instruction mapping module is used to output an anti-interference instruction set containing multiple instructions based on the main frequency band information and the interference feature information combined with preset instruction mapping rules.

[0082] The effect evaluation module is used to acquire the original mixed dataset and auxiliary sensing dataset within a preset acquisition period after the execution of the anti-interference instruction set, and use them as the verification mixed dataset and verification sensing dataset respectively, and combine the main frequency band information and the interference feature information to obtain the effect evaluation index of each instruction in the anti-interference instruction set;

[0083] The instruction adjustment module is used to compare each effect evaluation index with its preset effect threshold. If all effect evaluation indices reach their preset effect thresholds, the anti-interference instruction set remains unchanged. If at least one effect evaluation index fails to reach its preset effect threshold, the corresponding instruction in the anti-interference instruction set is adjusted until all effect evaluation indices reach their preset effect thresholds.

[0084] The feedback optimization module is used to obtain the adjustment frequency of different instructions within a preset optimization period, and update the preset step size when adjusting the instruction behavior according to the adjustment frequency of different instructions.

[0085] The core innovation of this embodiment lies in combining a multimodal sensor network with a closed-loop feedback mechanism in a dynamic and collaborative manner. This achieves real-time perception and adaptive suppression of environmental electromagnetic interference, resulting in a stable improvement in the signal-to-noise ratio of the target physical signal in an unshielded ordinary laboratory environment. Specifically, the data acquisition module deploys multimodal sensors to simultaneously acquire the target physical signal and environmental noise, avoiding the limitation of traditional single sensors in comprehensively sensing interference sources. The data processing module uses time-frequency analysis technology to extract structured interference features from the noise data, overcoming the shortcomings of static shielding methods in adapting to dynamic interference.

[0086] The instruction mapping module generates a joint anti-interference strategy based on preset rules, which includes physical shielding orientation adjustment, sampling timing window control, and digital filter parameter configuration, replacing the rigid processing method of fixed parameter filters. The effect evaluation module verifies the effect of the strategy execution through quantitative indicators, and the instruction adjustment module dynamically fine-tunes the instruction parameters according to the evaluation results, forming a complete feedback loop. The feedback optimization module adaptively updates the preset step size according to the frequency of instruction adjustment, ensuring that the system continuously optimizes the adjustment efficiency during long-term operation.

[0087] The aforementioned technical solution, through the collaborative work of multiple modules, transforms data acquisition from a passive recording to an active, optimized intelligent process. This effectively solves the problem of weak signal detection distortion caused by high-frequency intermittent electromagnetic interference from Wi-Fi, switching power supplies, and other sources in laboratory environments. It significantly reduces the stringent environmental requirements for high-quality experiments. Furthermore, by intelligently scheduling high-power sensors to operate only during periods when the data is valuable, it achieves a balance between system energy efficiency and data quality. Compared to existing technologies, this solution eliminates the need for expensive static shielding facilities, dynamically combating changing interference under ordinary laboratory conditions, and stably improving the signal-to-noise ratio of the target physical signal, thereby significantly enhancing the reliability and accuracy of physical experimental data.

[0088] In another embodiment, this application also provides a computer storage medium storing computer-executable instructions, which, when executed, implement the aforementioned method for acquiring physical experimental data based on multimodal sensors.

[0089] The above embodiments are only used to illustrate the technical methods of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of this application without departing from the spirit and scope of the technical methods of this application.

Claims

1. A method for acquiring physical experimental data based on multimodal sensors, characterized in that, Includes the following steps: Simultaneously acquire the original hybrid dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise. Spectral analysis is performed on the original mixed dataset to obtain the main frequency band information of the target physical signal, and time-frequency analysis is performed on the auxiliary sensing dataset to obtain interference characteristic information reflecting environmental electromagnetic interference. Based on the main frequency band information and the interference feature information, combined with the preset instruction mapping rules, an anti-interference instruction set containing multiple instructions is output; The original hybrid dataset and auxiliary sensing dataset within a preset acquisition period after the execution of the anti-interference instruction set are obtained, and these are used as the verification hybrid dataset and verification sensing dataset, respectively. The effectiveness evaluation index of each instruction in the anti-interference instruction set is obtained by combining the main frequency band information and the interference feature information. Each effect evaluation index is compared with its preset effect threshold. If each effect evaluation index reaches its preset effect threshold, the anti-interference instruction set remains unchanged. If at least one effect evaluation index fails to reach its preset effect threshold, adjust the corresponding instruction in the anti-interference instruction set until all effect evaluation indices reach their preset effect thresholds. Obtain the adjusted frequency of different instructions within the preset optimization period, and update the preset step size when adjusting the instruction behavior according to the adjusted frequency of different instructions.

2. The physical experiment data acquisition method based on multimodal sensors according to claim 1, characterized in that, The original mixed dataset is subjected to Fourier transform to obtain its power spectral density, and the continuous frequency range in the power spectral density that is greater than the first preset frequency threshold is taken as the main frequency band information of the target physical signal. Short-time Fourier transform is performed on the auxiliary sensing dataset to obtain time-spectrum data, and interference feature information including interference frequency band information, interference pulse information, and interference direction information is extracted based on the time-spectrum data.

3. The physical experiment data acquisition method based on multimodal sensors according to claim 2, characterized in that, The instruction mapping rule is a preset mapping function that takes main frequency band information and interference feature information as input data and different instructions as output data; The main frequency band information and the interference frequency band information are input into the instruction mapping rule to output the filter parameter instruction. The interference pulse information and the interference direction information are input into the instruction mapping rule to output the acquisition window instruction and the shielding position instruction respectively. The anti-interference instruction set is a collection of different instructions output based on the instruction mapping rule.

4. The physical experiment data acquisition method based on multimodal sensors according to claim 3, characterized in that, After controlling the physical shielding device to move to the corresponding position according to the shielding position command, the original mixed dataset and auxiliary sensing dataset under the corresponding sampling window within the preset acquisition period are obtained according to the acquisition window command. Then, the acquired original mixed dataset and auxiliary sensing dataset are filtered according to the filtering parameters in the filtering parameter command to obtain the verification mixed dataset and verification sensing dataset.

5. The physical experiment data acquisition method based on multimodal sensors according to claim 4, characterized in that, The frequency range overlapping between the main frequency band information and the interference frequency band information is used as the evaluation frequency band to obtain the original signal-to-noise ratio of the original hybrid dataset and the verification hybrid dataset within the evaluation frequency band. And verification signal-to-noise ratio Obtain the first effect evaluation index of the filter parameter command. ; Obtain the interference frequency band information and interference pulse information of the verification sensing dataset and use them as the verification frequency band information and verification pulse information, respectively, and use the frequency range corresponding to the verification frequency band information as the verification frequency band. The number of pulses in the verification sensor dataset within its sampling window is obtained based on the verification pulse information. The number of pulses in the auxiliary sensing dataset within its preset acquisition period is obtained based on the interference pulse information. The second effect evaluation index of the acquisition window command is obtained. , This corresponds to the time length of the sampling window. This corresponds to the time length of the preset data collection period; Obtain the signal power of the auxiliary sensing dataset within the verification frequency band. And obtain the signal power of the sensing dataset used for verification within the same verification frequency band. The third effect evaluation index of the shielding position command is obtained. .

6. The physical experiment data acquisition method based on multimodal sensors according to claim 5, characterized in that, The original signal-to-noise ratio The verification uses signal-to-noise ratio ,in, These represent the signal power and noise power of the original mixed dataset within the evaluation frequency band, respectively. These represent the signal power and noise power of the hybrid dataset used for verification within the evaluation frequency band, respectively.

7. The physical experiment data acquisition method based on multimodal sensors according to claim 5, characterized in that, Set corresponding first preset effect threshold, second preset effect threshold, and third preset effect threshold for the first effect evaluation index, the second effect evaluation index, and the third effect evaluation index, respectively. For the effect evaluation index that does not reach its preset effect threshold, according to the order of the shielding position instruction, sampling window instruction, and filtering parameter instruction, a preset step size is updated based on the parameters in the corresponding instruction. The preset step size includes the preset shielding step size, the preset sampling step size, and the preset filtering step size. Each update of any instruction is treated as an instruction adjustment action. The results of each instruction adjustment action are obtained, including the hybrid dataset for verification, the perception dataset for verification, and the effect evaluation index of each instruction, until the effect evaluation index of each instruction reaches its corresponding preset effect threshold.

8. The physical experiment data acquisition method based on multimodal sensors according to claim 7, characterized in that, The number of times the instruction adjustment behavior is performed based on all instructions within the preset optimization period. And the number of times a single instruction modifies behavior. To obtain the adjusted frequency of this single instruction And obtain the preset step size after the single instruction update. , This is the preset step size before the update of this single instruction. and These are the preset adjustment factor and the reference frequency, respectively, both ranging from (0, 1).

9. A physical experiment data acquisition module based on a multimodal sensor, characterized in that, Includes the following modules: The data acquisition module is used to synchronously acquire the original mixed dataset of the target physical signal under environmental electromagnetic interference within a preset acquisition period, as well as the auxiliary sensing dataset characterizing the environmental electromagnetic noise. The data processing module is used to perform spectral analysis on the original mixed dataset to obtain the main frequency band information of the target physical signal, and to perform time-frequency analysis on the auxiliary sensing dataset to obtain interference characteristic information reflecting environmental electromagnetic interference. The instruction mapping module is used to output an anti-interference instruction set containing multiple instructions based on the main frequency band information and the interference feature information combined with preset instruction mapping rules. The effect evaluation module is used to acquire the original mixed dataset and auxiliary sensing dataset within a preset acquisition period after the execution of the anti-interference instruction set, and use them as the verification mixed dataset and verification sensing dataset respectively, and combine the main frequency band information and the interference feature information to obtain the effect evaluation index of each instruction in the anti-interference instruction set; The instruction adjustment module is used to compare each effect evaluation index with its preset effect threshold. If all effect evaluation indices reach their preset effect thresholds, the anti-interference instruction set remains unchanged. If at least one effect evaluation index fails to reach its preset effect threshold, the corresponding instruction in the anti-interference instruction set is adjusted until all effect evaluation indices reach their preset effect thresholds. The feedback optimization module is used to obtain the adjustment frequency of different instructions within a preset optimization period, and update the preset step size when adjusting the instruction behavior according to the adjustment frequency of different instructions.

10. A computer storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed, they implement the physical experiment data acquisition method based on a multimodal sensor as described in any one of claims 1-8.

Images