A wireless instruction encryption method and system for a mine intrinsic safety device

Abstract

The application relates to the field of mine wireless communication safety, and discloses a wireless instruction encryption method and system for a mine intrinsic safety device. The method comprises the following steps: collecting historical mine environment data and extracting features to obtain a characteristic environment description; a nonlinear mapping function of multipath propagation characteristics and signal interference degree is constructed according to the characteristic environment description; current environment data is collected to calibrate the nonlinear mapping function, and a calibrated model is obtained; a prediction value of encryption strength is obtained based on the current environment data and the calibrated model; if the prediction value is lower than an encryption strength threshold value, a key parameter is adjusted to generate an optimized encryption configuration; the dynamic environment data is classified by using a support vector machine with the configuration as a reference, an interference area is divided, and an adjustment coefficient of the interference area is determined; the encryption parameter is adjusted in real time according to the coefficient to obtain an encryption strength adjustment scheme that matches actual requirements. The application solves the problem that the encryption strength of a wireless instruction does not match the actual requirements in a complex mine environment, and improves the safety and adaptability of wireless instruction transmission.

Description

Technical Field

[0001] This application relates to the field of wireless communication security technology in mines, and in particular to a wireless command encryption method and system for intrinsically safe mining equipment. Background Technology

[0002] In complex underground environments such as mine tunnels, wireless communication is a key technology for remote equipment control, data transmission, and personnel positioning. However, the mine environment is characterized by complex tunnel structures, variable geological conditions, and numerous electromechanical devices, which leads to severe multipath effects, signal attenuation, and strong electromagnetic noise interference in wireless signal propagation, posing a serious challenge to the reliability, real-time performance, and security of wireless command transmission.

[0003] Existing technologies for encryption of wireless communications in mines suffer from the following main drawbacks. First, conventional encryption strategies often employ static configurations, pre-setting fixed key lengths and encryption algorithm rounds, which cannot adapt to the dynamically changing signal quality and interference levels in the mining environment. For example, in high-attenuation areas such as tunnel corners and working faces, static high-strength encryption may lead to communication delays or even interruptions; while in areas with weaker interference, excessively high encryption strength results in unnecessary waste of computational resources and energy. Second, existing methods typically address signal attenuation or noise interference in isolation, lacking a comprehensive analysis of their coupled effects. In reality, multipath propagation and electromagnetic noise in mines are intertwined, jointly determining the ease with which signals can be intercepted and decrypted, and existing technologies have failed to establish a precise mapping model between environmental characteristics and the required encryption strength. Furthermore, existing solutions lack sufficient intelligence, relying heavily on manual experience for parameter adjustments, resulting in delayed responses and difficulty in meeting real-time, adaptive security requirements. Furthermore, some studies have attempted to introduce environmental awareness, but these only involve simple threshold judgments or linear adjustments, failing to fully utilize machine learning algorithms to extract deeper, non-linear propagation characteristics and interference patterns from historical and real-time data. This results in low precision in encryption strength adjustment and poor generalization ability. Finally, existing technologies lack consideration for the spatial heterogeneity of mines, failing to finely classify the entire communication area according to interference characteristics and implement differentiated parameter adjustments, thus failing to achieve globally optimal encryption resource configuration. These shortcomings collectively lead to the core technical problems of existing mining wireless communication systems in dynamic and complex environments: a mismatch between encryption strength and the environment, and an inability to balance security and efficiency.

[0004] To address the above shortcomings, this application combines sensor network environment perception, machine learning feature modeling, and adaptive parameter adjustment to construct a dynamic mapping relationship between environmental interference characteristics and encryption strength. It also realizes refined encryption strategy adjustment based on region classification, solving the problems of poor adaptability and security-efficiency imbalance of existing static encryption methods in complex dynamic environments of mines, and improving the security and reliability of wireless command transmission and system resource utilization. Summary of the Invention

[0005] This application provides a wireless command encryption method and system for intrinsically safe mining equipment, which solves the problems of poor adaptability and imbalance between security and efficiency of existing static encryption methods in complex dynamic environments of mines, and improves the security and reliability of wireless command transmission and the utilization rate of system resources.

[0006] In a first aspect, this application provides a wireless command encryption method for intrinsically safe mining equipment, the method comprising:

[0007] Step S101: Collect signal attenuation data and noise level data in the historical mine environment to obtain multi-dimensional historical environment data, and then use the support vector machine algorithm to extract features from the multi-dimensional historical environment data to obtain a characteristic environment description.

[0008] Step S102: Extract multipath propagation path features based on the characterized environment description, and construct a nonlinear mapping function between the multipath propagation path features and the degree of interference in signal transmission;

[0009] Step S103: Collect real-time signal attenuation data and noise level data in the current mine environment to obtain current environmental data. Based on the noise level data in the current environmental data, perform electromagnetic interference level calibration on the nonlinear mapping function to obtain the calibrated mapping model.

[0010] Step S104: Extract new multipath propagation path features from the current environment data, input the new multipath propagation path features into the calibrated mapping model, and obtain the predicted value of the encryption strength level under the current environment.

[0011] Step S105: Determine whether the predicted value is lower than the preset encryption strength threshold. If so, adjust the key parameters of the wireless command encryption of the intrinsically safe mining equipment to generate an optimized encryption configuration.

[0012] Step S106: Based on the optimized encryption configuration, the support vector machine algorithm is used to classify the dynamically changing signal attenuation data and noise level data in the mine environment, divide the interference area types, and determine the area-specific adjustment coefficient corresponding to any interference area type.

[0013] Step S107: Adjust the encryption parameters of the current intrinsically safe mining equipment wireless communication system in real time according to the specific adjustment coefficient of the region to obtain an encryption strength adjustment scheme that matches the actual needs of the mine.

[0014] Secondly, this application provides a wireless command encryption system for intrinsically safe mining equipment, used to implement the wireless command encryption method for intrinsically safe mining equipment, the system comprising:

[0015] The feature extraction module is used to collect signal attenuation data and noise level data in the historical mine environment through a sensor network to obtain multi-dimensional historical environmental data. Then, the support vector machine algorithm is used to extract features from the multi-dimensional historical environmental data to obtain a characteristic environmental description.

[0016] The mapping construction module is used to analyze the characteristics of multipath propagation paths based on the characteristic environment description, and construct a nonlinear mapping function between the multipath propagation path characteristics and the degree of interference in signal transmission;

[0017] The model calibration module is used to collect real-time signal attenuation data and noise level data in the current mine environment, obtain current environmental data, and perform electromagnetic interference level calibration on the nonlinear mapping function based on the noise level data in the current environmental data to obtain the calibrated mapping model.

[0018] The strength prediction module is used to extract new multipath propagation path features from the current environment data, input the new multipath propagation path features into the calibrated mapping model, and obtain a predicted value of the encryption strength level under the current environment.

[0019] The encryption configuration module is used to determine whether the predicted value is lower than the preset encryption strength threshold. If so, it adjusts the key parameters of the wireless command encryption of the intrinsically safe mining equipment to generate an optimized encryption configuration.

[0020] The region classification module is used to classify the dynamically changing signal attenuation data and noise level data in the mine environment based on the optimized encryption configuration and the support vector machine algorithm, to divide the interference region types and determine the region-specific adjustment coefficient corresponding to any interference region type.

[0021] The parameter adjustment module is used to adjust the encryption parameters of the current intrinsically safe mining equipment wireless communication system in real time according to the specific adjustment coefficient of the region, so as to obtain an encryption strength adjustment scheme that matches the actual needs of the mine.

[0022] This application proposes a wireless command encryption method and system for intrinsically safe mining equipment, solving the problems of poor adaptability and security-efficiency imbalance of existing static encryption methods in the complex dynamic environment of mines, and improving the security and reliability of wireless command transmission and system resource utilization. Compared with the prior art, the beneficial effects of the technical solution of this application are at least as follows:

[0023] First, by collecting and extracting features from historical and real-time environmental data of the mine, and constructing a nonlinear mapping model of multipath propagation characteristics and interference degree, we have achieved quantitative characterization and accurate prediction of environmental interference characteristics, overcoming the shortcomings of traditional methods that rely on fixed thresholds and cannot reflect the complex nonlinearity of the environment.

[0024] Second, based on the calibrated environment mapping model, the encryption strength requirement is predicted, and the key parameters are dynamically adjusted by judging the difference between the predicted value and the threshold to generate an optimized encryption configuration. This enables the encryption strength to adapt to environmental changes, ensuring communication security while avoiding resource waste. It solves the problems of insufficient static configuration under high interference and excessive configuration under low interference.

[0025] Third, based on optimized configuration, the dynamically changing environmental data is classified, different types of interference areas are identified, and specific adjustment coefficients for each area are determined. This enables refined perception of the spatial heterogeneity of the mine and the deployment of differentiated encryption strategies, improving the targeting of encryption parameter adjustments and the overall system's adaptability.

[0026] Fourth, the encryption parameters of the wireless communication system are modified in real time according to the specific adjustment coefficient of the region, forming an encryption strength adjustment scheme that matches the actual needs. This realizes the fully automated closed-loop adjustment from environmental perception, model prediction, parameter optimization to regional adaptation, improving the system's response speed, safety and reliability, and resource utilization efficiency in complex and ever-changing mining environments. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart illustrating a wireless command encryption method for an intrinsically safe mining device according to this application.

[0029] Figure 2 Here is a diagram of the neural network model structure for constructing the nonlinear mapping function in this application;

[0030] Figure 3 This is a comprehensive comparison of the performance of the nonlinear mapping function before and after calibration in this application;

[0031] Figure 4 The comparison results are as follows: (This refers to the cumulative distribution function of the prediction error of the degree of interference in this application.)

[0032] Figure 5 This is a schematic diagram of the structure of a wireless command encryption system for intrinsically safe mining equipment according to this application. Detailed Implementation

[0033] This application provides a wireless command encryption method and system for intrinsically safe mining equipment. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0034] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of a wireless command encryption method for intrinsically safe mining equipment in this application includes:

[0035] Step S101: Collect signal attenuation data and noise level data in the historical mine environment to obtain multi-dimensional historical environmental data. Then, use the support vector machine algorithm to extract features from the multi-dimensional historical environmental data to obtain a characteristic environmental description.

[0036] In one specific embodiment, step S101 may specifically include the following steps:

[0037] The signal attenuation data and noise level data in the historical mine environment are collected by a sensor network. The signal attenuation data characterizes the energy loss during wireless command transmission, and the noise level data characterizes the interference intensity of the mine environment on wireless command transmission.

[0038] Filtering and normalizing signal attenuation data and noise level data from historical mining environments are performed to generate multidimensional historical environmental data.

[0039] The support vector machine algorithm is used to extract features from multidimensional historical environmental data to obtain feature vectors.

[0040] Characteristic environment descriptions are generated based on feature vectors. These descriptions contain feature information about signal propagation characteristics and noise distribution.

[0041] Specifically, in mining operations, a sensor network consisting of multiple distributed sensor nodes is deployed. These nodes are evenly distributed across different locations within the mine roadways, covering areas with varying geological structures and depths. The network collects real-time signal attenuation and noise level data from the historical mining environment. Signal attenuation data is obtained by sending test wireless commands from the sensor nodes. The receiving end records the signal strength difference before and after the command transmission, expressed in decibels (dB), representing the energy loss during wireless command transmission. For example, in areas with thick coal seams, the signal may attenuate by 0.8 dB per meter, while in rock roadways, it may attenuate by 1.5 dB per meter. Noise level data is obtained by detecting the intensity of interference signals such as electromagnetic interference and mechanical vibration in the mining environment. Power spectral density (PSD) is used as a quantitative indicator to represent the interference intensity of the mining environment on wireless command transmission. For instance, the background noise PSD in shallow mining areas may be -90 dBm / Hz, while in deeper areas affected by equipment vibration, it may reach -75 dBm / Hz. During the data collection process, the sensor nodes collect data every 10 seconds, continuously for one month to form a historical dataset, ensuring that the data covers various changes in the mining environment.

[0042] The collected historical signal attenuation data and noise level data are preprocessed. First, mean filtering is used to remove sudden noise from the data. A sliding window size of 5 is set, and the average value of the data within the window is used to replace the center data point. For example, for the signal attenuation data sequence [0.7, 0.9, 2.1, 0.8, 0.6], mean filtering transforms it into [0.7, 0.8, 0.9, 0.8, 0.6], effectively smoothing data fluctuations. Then, normalization is performed using a minimum-maximum normalization method. The historical acquisition range of the signal attenuation data is statistically analyzed. Through formula Attenuate the original signal data Convert to standardized data in the range [0,1] For noise level data, the historical collection range is: Through formula Original noise level data Convert to standardized data in the range [0,1] Standardized signal attenuation data and noise level data Combined, multi-dimensional historical environmental data is formed, with each data sample... The vector form of the data ensures the uniformity of the dimensions of the data across different dimensions, laying the foundation for subsequent algorithm processing.

[0043] The Support Vector Machine (SVM) algorithm is used to extract features from multidimensional historical environmental data. The kernel function of the algorithm is set as the radial basis function, and the kernel function parameters are... The value is set to 0.5, the penalty coefficient is set to 1.0, and the slack variable is set to 0.1. After inputting the multidimensional historical environment data into the support vector machine algorithm, the algorithm first calculates the similarity between different data samples, and then maps the multidimensional historical environment data to a high-dimensional feature space through the radial basis function, where the standardized signal attenuation data... The mapped data dimensions primarily reflect the energy loss patterns in wireless command transmission, while the standardized noise level data... The mapped data dimensions mainly reflect the distribution characteristics of environmental interference. In the high-dimensional feature space, the algorithm finds the optimal hyperplane by maximizing the classification margin, selects support vectors that can distinguish the characteristics of different mine environments, and then extracts feature vectors that characterize signal propagation characteristics and noise distribution patterns. Each element of the feature vector corresponds to a feature dimension in the high-dimensional space. For example, in the feature vector [0.32, 0.67], the first element characterizes the key feature of signal attenuation under this environment, and the second element characterizes the core feature of noise distribution.

[0044] Based on the extracted feature vectors, a characteristic environment description is generated. The characteristic environment description is presented in the form of a set of vectors. Each feature vector corresponds to a set of feature information of a specific mine environment, including signal propagation characteristics and noise distribution features. For example, the characteristic environment description corresponding to a certain feature vector is "signal attenuation gradient 0.32, noise distribution variance 0.67", where the signal attenuation gradient quantifies the energy loss rate of signal transmission in this environment, and the noise distribution variance reflects the stability of environmental interference.

[0045] This step collects real historical environmental data through a sensor network. After preprocessing, the data noise and dimensional differences are eliminated. The support vector machine algorithm accurately extracts environmental features, and the generated characteristic environmental description can quantify the impact of different mine environments on wireless command transmission. This provides data support for the subsequent construction of mapping relationships and dynamic adjustment of encryption parameters. It solves the technical problem in traditional methods where environmental features cannot be effectively quantified, resulting in a lack of basis for adjusting encryption parameters. This enables the encryption configuration to meet the actual environmental needs of the mine.

[0046] Step S102: Extract multipath propagation path features based on the characteristic environment description, and construct a nonlinear mapping function between multipath propagation path features and the degree of interference in signal transmission.

[0047] In one specific embodiment, step S102 may specifically include the following steps:

[0048] Extract multipath propagation path features based on the characteristic environment description, and obtain the number and intensity of multipath propagation paths;

[0049] If the number of multipath propagation paths exceeds a preset path threshold, a neural network model is used to construct a nonlinear mapping function between the characteristics of multipath propagation paths and the degree of interference in signal transmission.

[0050] Specifically, based on the generated characteristic environment description, signal propagation characteristics and noise distribution features are analyzed to extract multipath propagation path features. The number of multipath propagation paths is determined by the number of inflection points in the signal attenuation gradient. An inflection point is defined as the difference between adjacent signal attenuation gradients exceeding 0.15, and each inflection point corresponds to a multipath propagation path. The strength of each path is calculated by combining the noise distribution variance. The calculation formula is If the noise distribution variance is 0.25 and the mean signal attenuation gradient is 0.36 in a certain set of characteristic environment descriptions, then the corresponding path strength is... This ultimately forms a number of paths. and the strength of each path are Multipath propagation path feature dataset.

[0051] The preset path threshold is set based on the geological structure and depth of the mine environment. In shallow mine areas (depth ≤ 600 meters), the geological structure is simple, and the impact of multipath propagation is weak; the preset path threshold is set to 3. In deep mine areas (depth > 600 meters), due to factors such as rock reflection and equipment obstruction, multipath propagation is prominent; the preset path threshold is set to 5. The number of extracted multipath propagation paths is then calculated. Compare with the preset path threshold for the corresponding area; if the number of paths... If the number of paths does not exceed the preset path threshold, it indicates that the multipath interference in the environment is mild, and the default mapping relationship will be used directly; if the number of paths... If the number of paths extracted from the deep mining area exceeds the preset path threshold, such as 6 (exceeding the threshold of 5), a neural network model is used to construct a nonlinear mapping function between the multipath propagation path characteristics and the degree of interference in signal transmission.

[0052] The neural network model uses a multilayer perceptron structure; please refer to [link / reference]. Figure 2 It consists of an input layer, hidden layers, and an output layer. The number of neurons in the input layer is consistent with the dimension of the multipath propagation path features. If the multipath propagation path features include the number of paths *m* and the strengths of four fixed main paths... The input layer directly corresponds to the 5-dimensional input features, with 5 neurons. Each neuron receives feature data from one dimension, and the input vector is... The hidden layer uses a single-layer structure, with the number of neurons calculated using the formula "Number of neurons in the hidden layer = 2 × Number of neurons in the input layer + 3", which is set to 13. The weights of the hidden layer are initialized using a He normal distribution, and the bias term is initialized to 0.01. The hidden layer uses the ReLU activation function. The number of neurons in the output layer is set to 1, and the output value is the degree of interference in signal transmission (range 0~1). The output layer uses the Sigmoid function to ensure that the output result is within a reasonable range. All layers of the model are connected using a fully connected method. The connection weight matrix between the input layer and the hidden layer is set to A (dimension 5×13), and the weight matrix between the hidden layer and the output layer is set to B (dimension 13×1). Each neuron in the hidden layer corresponds to one bias term b.

[0053] During neural network model training, the training dataset consists of historical multipath propagation path feature data and corresponding measured data on signal transmission interference levels. The historical multipath propagation path feature data comes from the analytical results of past characteristic environment descriptions. The measured data on signal transmission interference levels is obtained through signal quality detection equipment deployed in different areas of the mine. For example, when the number of multipath propagation paths is 6, and the strengths of the four main paths are fixed at 0.27, 0.31, 0.24, and 0.29 respectively, the measured signal transmission interference level is 0.78, forming a set of training samples. The training process uses the Adam optimizer, with a learning rate of 0.002, a batch size of 64, and 1000 iterations. The loss function is... Use the mean square error function, i.e. (in These are the model's predicted values. This represents the measured value of the degree of interference in signal transmission. (The number of samples is used as the basis for the algorithm). The network weights and biases are continuously adjusted through backpropagation until the loss function value converges to below 0.004, completing the model training. The trained neural network model is a nonlinear mapping function. By inputting the multipath propagation path features extracted in real time into this function, the corresponding degree of interference in signal transmission can be output.

[0054] By parsing the characteristic environment description to obtain multipath propagation path features, and combining them with a preset threshold to determine whether to build a neural network model, the trained model is used to establish a precise mapping relationship between multipath propagation and interference level. This solves the technical problem of difficulty in quantifying multipath propagation interference, provides an accurate basis for interference assessment for subsequent dynamic adjustment of encryption parameters, and enables precise matching between encryption strength and signal transmission interference.

[0055] Step S103: Collect real-time signal attenuation data and noise level data in the current mine environment to obtain current environmental data. Based on the noise level data in the current environmental data, perform electromagnetic interference level calibration on the nonlinear mapping function to obtain the calibrated mapping model.

[0056] In one specific embodiment, step S103 may specifically include the following steps:

[0057] The current environmental data is obtained by collecting real-time signal attenuation data and noise level data in the current mine environment through a sensor network, and then filtering and normalizing the data.

[0058] Using the actual electromagnetic interference impact corresponding to the noise level data in the current environmental data as a reference, the deviation between the output value of the nonlinear mapping function and the actual electromagnetic interference impact is calculated;

[0059] Based on the bias, the weights of the nonlinear mapping function are adjusted using the minimum mean square error method to generate a calibrated mapping model that characterizes the environmental disturbance properties.

[0060] Specifically, sensor network nodes are distributed in key areas of the mine roadways, deployed at fixed intervals to achieve full coverage, and synchronously collect signal attenuation data and noise level data in the current environment. Signal attenuation data is obtained by transmitting test wireless commands at a specific frequency between nodes, recording the power difference before and after transmission, in decibels; noise level data is collected by the built-in detection module of the sensors, covering the power spectral density of interference signals such as electromagnetic interference and mechanical vibration, in dBm / Hz, with the acquisition frequency set to twice per second, continuously collecting 10 sets of data to form the raw dataset, avoiding random errors from single acquisitions.

[0061] The raw data processing follows the previously established standard procedure. First, a moving average filtering algorithm is applied, with a moving window size of 3, to smooth both signal attenuation and noise level data, eliminating fluctuations caused by sudden interference and random noise, and ensuring data stability. After filtering, normalization is performed using a minimum-maximum normalization method, with the value range determined by referencing historical mine data. and Through formula Filtered data Convert to standardized data in the [0,1] interval ; through formula Filtered data Convert to standardized data The current environmental data is formed by combining two types of standardized data, in the following format: This ensures that the data units are consistent and that the data format requirements for subsequent calculations are met.

[0062] Based on the standardized noise level data in the current environmental data The corresponding actual electromagnetic interference impact is used as a reference. The actual electromagnetic interference impact is obtained through a signal quality detector deployed next to the same sensor node. This instrument directly measures the bit error rate of wireless command transmission in the current environment and converts the bit error rate into a quantized value ranging from [0,1]. t Bit error rate and t The values ​​are positively correlated; for example, a bit error rate of 0.02 corresponds to... t =0.3. The current environmental data is input into a nonlinear mapping function, and the output is a predicted value of the degree of interference in signal transmission. Deviation is expressed by the formula Calculate, if t =0.3、 =0.25, then Δt=0.05. This value directly reflects the prediction deviation of the mapping function under the current environment, providing a quantitative basis for subsequent weight adjustment.

[0063] Based on the calculated deviation Before adjusting the weights of the nonlinear mapping function using the minimum mean square error method, it is necessary to clarify that the neural network model of the nonlinear mapping function is a three-layer perceptron structure, including an input layer, a hidden layer, and an output layer. If the number of neurons in the input layer is 5, the number of neurons in the hidden layer is set to 13, and the number of neurons in the output layer is 1, the ReLU activation function is used in the hidden layer, and the Sigmoid activation function is used in the output layer. The connection weight matrix from the input layer to the hidden layer is set to A (dimension 5×13), and the connection weight vector from the hidden layer to the output layer is set to B (dimension 13×1). Each neuron in the hidden layer corresponds to one bias term b. With the goal of minimizing the mean square error, the mean square error function is defined as... The partial derivative of the mean squared error with respect to weight B, calculated using the chain rule, is: ,in For the first i The output value of each hidden layer neuron For hidden layers to the 1st i The connection weight vector of each hidden layer neuron.

[0064] With a preset learning rate η=0.002 (set based on the fluctuation characteristics of mine environment data to avoid weight oscillations or slow convergence), the weight update formula is as follows: Taking specific data as an example, if Output value of a neuron in a hidden layer =0.4, initial weight =0.3, first calculate the partial derivative. Then substitute into the update formula to get The above calculations are performed on each of the 13 hidden layer neurons one by one, updating the connection weights from all hidden layers to the output layer, resulting in a new weight vector. (Dimension 13×1). Repeat the above weight adjustment process, recalculating the deviation between the predicted value and the actual electromagnetic interference impact after each adjustment, until the deviation is calculated for 5 consecutive times. All values ​​are less than 0.015. At this point, the prediction accuracy of the mapping function can meet the needs of practical applications, and a calibrated mapping model is generated.

[0065] By calculating the deviation with reference to the actual electromagnetic interference, the prediction error caused by environmental changes is accurately captured. Then, the function weights are dynamically adjusted by the minimum mean square error method, so that the mapping model can fit the current environmental interference characteristics in real time, effectively solving the prediction deviation problem caused by the dynamic changes of the environment.

[0066] Step S104: Extract new multipath propagation path features from the current environment data, input the new multipath propagation path features into the calibrated mapping model, and obtain the predicted value of the encryption strength level under the current environment.

[0067] In one specific embodiment, step S104 may specifically include the following steps:

[0068] The support vector machine algorithm is used to extract features from the current environmental data, obtain the current environmental feature vector, and extract new multipath propagation path features from it;

[0069] The new multipath propagation path characteristics are input into the calibrated mapping model, and the new multipath propagation path characteristics are calculated through forward propagation to output the degree of interference to signal transmission in the current environment.

[0070] The degree of interference is converted into a predicted value of the encryption strength level under the current environment. The predicted value represents the security strength level required for wireless command encryption in the current mining environment.

[0071] Specifically, the current environmental data after filtering and normalization is... It exists in vector form. Corresponding to the standardized signal attenuation data, The standardized noise level data, along with the standardized noise level data, together reflect the comprehensive impact of the current mine environment on wireless command transmission. When using the support vector machine algorithm to extract features from the current environmental data, the kernel function is set to a radial basis function, with a kernel parameter value of 0.6, a penalty coefficient of 1.2, and a slack variable of 0.08. These parameters are determined based on the distribution characteristics of the mine environmental data and the feature extraction target. The current environmental data is then processed using... After the data is input into the Support Vector Machine (SVM) algorithm in vector form, the algorithm maps the two-dimensional data to a high-dimensional feature space through the radial basis function. It then filters the support vectors by maximizing the classification margin and extracts the current environment feature vector based on the support vectors. The dimension of this vector is consistent with the dimension of the high-dimensional space, and each element corresponds to a feature dimension in the high-dimensional space that is related to the propagation or interference of environmental signals.

[0072] When extracting new multipath propagation path features from the current environmental feature vector, the number of multipath propagation paths is determined by analyzing the components related to signal attenuation gradient in the feature vector. A path inflection point is defined when the difference between adjacent feature components exceeds 0.15, and each inflection point corresponds to one multipath propagation path. The strength of each current path is calculated by combining the noise distribution-related components in the feature vector. The calculation formula is ,in, Let be the variance of the noise distribution in the current environmental feature vector, and g be the mean of the current environmental signal attenuation gradient. For example, if the variance of the noise distribution in the current environmental feature vector is 0.28, and the mean of the current environmental signal attenuation gradient is 0.35, then the path strength... This ultimately results in a number of paths Q and four fixed main path strengths. The new multipath propagation path features, the feature vector form is as follows: It has 5 dimensions, which matches the input dimensions of the calibrated mapping model.

[0073] The calibrated mapping model is a three-layer perceptron neural network. The input layer has 5 neurons, receiving data from the five dimensions of the new multipath propagation path features. The number of hidden layer neurons is calculated using the formula "Number of hidden layer neurons = 2 × Number of input layer neurons + 3", resulting in 13 neurons. Hidden layer weights are initialized using a He normal distribution, with a bias term initialized to 0.01, and the ReLU activation function is used. The output layer has 1 neuron, and the Sigmoid activation function is used. Both the input and hidden layers, and the hidden and output layers, are fully connected. The connection weight matrix from the input layer to the hidden layer is set to C (5×13 dimensions), and the connection weight vector from the hidden layer to the output layer is set to D (13×1 dimensions). After inputting the new multipath propagation path features, calculations are performed through forward propagation. Input layer neurons transmit data from each dimension to the hidden layer. Each hidden layer neuron receives the sum of the product of the corresponding input layer data and the corresponding element of the weight matrix C. After adding the bias term, the model is activated by the ReLU function to obtain the hidden layer output value. The outputs of the 13 hidden layer neurons form the hidden layer output vector (13×1 dimensions). The hidden layer output vector is multiplied by the corresponding elements of the weight vector D and then summed. After activation by the Sigmoid function, the output shows the degree of interference to signal transmission in the current environment, R. The value of R is in the range of [0,1]. The larger the value, the stronger the interference to signal transmission in the environment. For example, when the output R=0.68, it indicates that the current environmental interference is at a moderate to strong level.

[0074] The interference level R is converted into a predicted encryption strength level H under the current environment using a linear mapping formula. ,in and The conversion coefficient is set based on the safety requirements of different areas of the mine, for the deep areas of the mine. =15, =3, Middle layer area of ​​the mine =12, =2.5, shallow area of ​​the mine =10, =2. For example, in deep mining areas, if =0.6, then In the shallow areas of the mine, if =0.2, then The predicted value H ranges from 2 to 18, and the higher or lower the value, the higher the security level required for wireless command encryption.

[0075] By comprehensively capturing key information related to signal attenuation, noise, their coupling, and multipath propagation in the current environment through five-dimensional feature vectors, this approach overcomes the shortcomings of existing technologies that isolate and process single environmental factors. The support vector machine algorithm's extraction logic for five-dimensional data ensures the integrity and relevance of feature information. The forward propagation calculation of the neural network model establishes a precise correspondence between feature data and the degree of interference. Linear mapping enables an effective transformation from the degree of interference to the predicted encryption strength value, allowing the encryption strength prediction to fully fit the complex nonlinear characteristics of the mine environment. This avoids the mismatch between static encryption configuration and dynamic environment, and improves the adaptability and targeting of the encryption strategy.

[0076] Step S105: Determine whether the predicted value is lower than the preset encryption strength threshold. If so, adjust the key parameters of the wireless command encryption for the intrinsically safe mining equipment to generate an optimized encryption configuration.

[0077] In one specific embodiment, step S105 may specifically include the following steps:

[0078] Set an encryption strength threshold to determine whether the current encryption strength meets the data transmission security requirements in a mining environment;

[0079] Determine if the predicted value is lower than the encryption strength threshold. If so, initiate the key parameter adjustment process for wireless command encryption of intrinsically safe mining equipment. The key parameters include key length and number of encryption rounds.

[0080] Based on the difference between the predicted value and the encryption strength threshold, the adjustment direction and adjustment step size of the key parameters are determined.

[0081] Based on the determined adjustment direction and adjustment step size, the key parameters are iteratively optimized using the gradient descent algorithm;

[0082] Determine whether the encryption strength corresponding to the optimized key parameters exceeds the encryption strength threshold. If so, generate an optimized encryption configuration based on the optimized key parameters.

[0083] Specifically, the encryption strength threshold is set based on the geological structure, equipment distribution density, and communication security level requirements of different areas of the mine. In deep areas of the mine (depth > 600 meters), which are significantly affected by multipath effects and electromagnetic interference, the threshold is set to 12; in mid-level areas (depth 300-600 meters), the interference level is moderate, and the threshold is set to 10; in shallow areas (depth ≤ 300 meters), the interference is weak, and the threshold is set to 8. The threshold is stored in a system preset module and serves as a benchmark for determining whether the current encryption strength meets the data transmission security requirements. Its value and the predicted encryption strength value use the same quantification standard to ensure consistency in the judgment logic. The predicted encryption strength value obtained in step S104 is compared with the encryption strength threshold for the corresponding area. If the predicted value is lower than the threshold, it indicates that the current encryption strength cannot withstand the security risks caused by environmental interference, and the key parameter adjustment process is initiated. Key parameters include key length and number of encryption rounds. The initial key length is set to 128 bits, and the initial number of encryption rounds is set to 10 rounds. Both are core parameters affecting encryption strength. The longer the key length and the more encryption rounds, the higher the encryption strength, but this also increases the equipment's computational load and communication latency.

[0084] The adjustment direction and step size are determined based on the difference between the predicted value and the encryption strength threshold. The formula for calculating the difference is as follows: ,in This is the encryption strength threshold. This represents the predicted encryption strength level under the current environment. A positive difference indicates that the encryption strength needs to be increased, by increasing the key length and the number of encryption rounds; a negative difference indicates no adjustment is needed. The adjustment step size is positively correlated with the difference, and the step size calculation formula is: Key Length Adjustment Step Size The number of encryption rounds is adjusted by step size. Where `round` is the rounding function. For example, in deep mining areas, the encryption strength threshold... =12, if the predicted value of the encryption strength level under the current environment is... =9, the difference between the two =3, then the key length adjustment step size Bit, encryption round number adjustment step size Round; if the predicted value of the encryption strength level under the current environment =14, difference =-2, then the key length adjustment step size Bit, encryption round number adjustment step size wheel.

[0085] The gradient descent algorithm is used to iteratively optimize the key parameters. The learning rate ζ = 0.15, the maximum number of iterations is set to 20, and the convergence condition is that the difference between the optimized encryption strength and the threshold is less than 0.5. The algorithm constructs a loss function with key length and number of encryption rounds as optimization variables and achieving the encryption strength threshold as the objective. The loss function expression is as follows: ,in This represents the predicted encryption strength value corresponding to the optimized key parameters. During the iterative optimization process, the parameter values ​​are adjusted in each round based on the partial derivative of the loss function with respect to the key parameters. The key length update formula is as follows: The formula for updating the number of encryption rounds is: ,in Let n be the key length and the number of encryption rounds in the nth iteration. Let be the parameter value for the (n+1)th iteration, and exp be the exponential function. For example, the initial key length. Bits, number of encryption rounds Wheel, difference =3, key length adjustment step size =24 bits, encryption round number adjustment step size In the first iteration of the 6th round, if the loss function is... ,but Bit, Then, calculate the predicted encryption strength value corresponding to this parameter. Substitute into the loss function to update Then, proceed to the next iteration.

[0086] After each iteration, the predicted encryption strength value corresponding to the current key parameters is calculated using the encryption strength evaluation model. The evaluation model takes key length and number of encryption rounds as input and outputs a quantified encryption strength value. The mapping relationship is obtained by fitting a large amount of encryption test data; for example, a 140-bit key length and 13 encryption rounds correspond to an encryption strength of 11.2, while a 150-bit key length and 15 encryption rounds correspond to an encryption strength of 12.5. The predicted encryption strength value is then determined. Does it exceed the encryption strength threshold? If the limit is exceeded, the iteration stops, and the current key parameters are determined as the optimized parameters; if the limit is not exceeded and the number of iterations has not reached the upper limit, the iteration optimization continues. When the condition is still not met after 20 iterations, the parameters corresponding to the upper limit of iterations are used as the optimization result to ensure that the process terminates normally.

[0087] An optimized encryption configuration is generated based on the optimized key parameters. The configuration file includes the key length, number of encryption rounds, and corresponding encryption algorithm parameters, and is stored in the encryption module of the intrinsically safe mining equipment as the execution standard for subsequent wireless command encryption. This process dynamically adjusts the key parameters, enabling the encryption strength to adaptively match environmental interference, thus solving the problem that existing static encryption configurations cannot adapt to the dynamic environment of mines. While ensuring communication security, by reasonably setting the adjustment step size and iterative convergence conditions, it avoids resource waste and communication delays caused by excessive adjustment of encryption strength, thereby improving the system's adaptability and operating efficiency.

[0088] Step S106: Based on the optimized encryption configuration, the support vector machine algorithm is used to classify the dynamically changing signal attenuation data and noise level data in the mine environment, divide the interference area types, and determine the region-specific adjustment coefficient corresponding to any interference area type.

[0089] In one specific embodiment, step S106 may specifically include the following steps:

[0090] The dynamic environmental feature vector is obtained by collecting dynamic signal attenuation data and noise level data in the mine environment in real time through sensor network and normalizing the data.

[0091] Based on the baseline security reference corresponding to the optimized encryption configuration, the support vector machine algorithm is used to classify the dynamic environment feature vectors and divide them into different types of interference regions.

[0092] For any type of interference area, based on the dynamic environmental feature vector corresponding to that area type and the optimized encryption configuration, the adjustment range of the wireless command encryption parameters under that area type is calculated, and the area-specific adjustment coefficient corresponding to any type of interference area is determined.

[0093] Specifically, real-time acquisition of dynamically changing signal attenuation and noise level data in the mine environment, followed by normalization to obtain a dynamic environment feature vector, is performed. The specific acquisition method and processing procedure are consistent with steps S101 and S103, and will not be repeated here. The standardized signal attenuation and noise level data are then combined sequentially to form a 2-dimensional dynamic environment feature vector, achieving dimension unification of data from different dimensions and providing standardized input for subsequent algorithm processing.

[0094] Optimized encryption configuration includes key length and encryption rounds The corresponding benchmark security reference is the anti-cracking capability and communication efficiency threshold of this configuration under standard test environment. The anti-cracking capability is quantified by the key space size, and the communication efficiency threshold is defined by the upper limit of transmission delay. Based on this benchmark security reference, the support vector machine algorithm is used to classify the dynamic environment feature vectors. The kernel function of the algorithm is the radial basis function, with a kernel function parameter of 0.7, a penalty coefficient of 2.0, and a slack variable of 0.15. The algorithm first maps the dynamic environment feature vectors to a high-dimensional feature space, and then determines the optimal classification hyperplane by maximizing the classification interval between samples of different categories. Based on the relative position and distance between the dynamic environment feature vectors and the hyperplane, the mine communication area is divided into three types of interference areas: high interference area, medium interference area, and low interference area. The classification judgment rule is implemented through the boundary threshold determined during the algorithm training process. If the distance from the dynamic environment feature vector to the hyperplane is greater than 0.5, it is judged as a high interference area; if the distance is less than -0.5, it is judged as a low interference area; and if the distance is between [-0.5, 0.5], it is judged as a medium interference area.

[0095] For any type of interference region, a calculation model for the adjustment range of encryption parameters is constructed based on the dynamic environmental feature vector corresponding to that region and the optimized encryption configuration. The input to the model is the dynamic environmental feature vector. Optimize key length Optimize the number of encryption rounds The output is the key length adjustment range. Adjustment range of encryption rounds The formula for calculating the key length adjustment range is as follows: ,in The weighting coefficients represent the degree of influence of signal attenuation data and noise level data on key length adjustment, respectively; the formula for calculating the adjustment range of encryption rounds is... ,in These are weighting coefficients, representing the equal impact of signal attenuation data and noise level data on the adjustment of the number of encryption rounds. Region-specific adjustment coefficients include key length adjustment coefficients. And encryption round adjustment coefficient ,in , This calculation method establishes a direct correlation between dynamic environmental characteristics and encryption parameter adjustment coefficients.

[0096] This process captures dynamic environmental data from the mine in real time through a sensor network. After standardization, interference caused by data differences is eliminated. The support vector machine algorithm, based on a baseline safety reference, achieves accurate classification of interference areas, solving the problem of existing technologies lacking consideration for the spatial heterogeneity of mines. The construction of region-specific adjustment coefficients establishes a quantitative relationship between dynamic environmental characteristics and encryption parameter adjustments, providing a clear basis for adjusting encryption parameters in different interference areas. This avoids the problem of insufficient security in high-interference areas and resource waste in low-interference areas caused by uniform encryption configuration, improving the adaptability and targeting of encryption strategies to the complex dynamic environment of mines.

[0097] Step S107: Adjust the encryption parameters of the current intrinsically safe mining equipment wireless communication system in real time according to the region-specific adjustment coefficient to obtain an encryption strength adjustment scheme that matches the actual needs of the mine.

[0098] In one specific embodiment, step S107 may specifically include the following steps:

[0099] Obtain the region-specific adjustment coefficient and the encryption parameters of the current intrinsically safe mining equipment wireless communication system. The encryption parameters include the key length and the number of encryption rounds.

[0100] Using region-specific adjustment coefficients as weights, the encryption parameters are weighted and calculated to obtain the initial encryption parameters;

[0101] Determine whether the encryption strength corresponding to the initial encryption parameters matches the security requirements of the current mine interference area type. If not, iteratively adjust the encryption parameters based on the area-specific adjustment coefficient until the encryption strength meets the security requirements of the current mine interference area type.

[0102] The encryption parameters that meet the security requirements are determined as the final encryption parameters, and an encryption strength adjustment scheme that matches the actual needs of the mine is generated based on the final encryption parameters.

[0103] Specifically, the region-specific adjustment coefficients obtained in step S106 include a key length adjustment coefficient and an encryption round number adjustment coefficient. Both are quantized coefficients calculated based on dynamic environment feature vectors, with values ​​ranging from 0.3 to 1.8, corresponding to the adjustment weights of the key length and encryption round number, respectively. The encryption parameters of the current intrinsically safe mining equipment wireless communication system are read from the encryption configuration module, including the current key length and the current encryption round number. The current key length is the key length determined in the optimized encryption configuration, and the current encryption round number is the encryption round number determined in the optimized encryption configuration. These two types of parameters correspond one-to-one with the region-specific adjustment coefficients. The key length adjustment coefficient is specifically used to adjust the current key length, and the encryption round number adjustment coefficient is specifically used to adjust the current encryption round number.

[0104] Using a region-specific adjustment coefficient as weight, the corresponding encryption parameters are weighted and calculated to obtain the initial encryption parameters. The weighted calculation adopts the mathematical logic of direct multiplication. The initial key length is the result of multiplying the current key length by the key length adjustment coefficient, and the initial number of encryption rounds is the result of multiplying the current number of encryption rounds by the number of encryption rounds adjustment coefficient. After the calculation is completed, the result is rounded to the nearest integer to ensure that the parameters meet the execution standards of the device's encryption module. For example, in high-interference regions, the key length adjustment factor is 1.72 and the encryption round adjustment factor is 1.86. If the current key length is 144 bits and the current encryption round is 14, then the initial key length is 144 multiplied by 1.72 and rounded to 248 bits, and the initial encryption round is 14 multiplied by 1.86 and rounded to 26. In low-interference regions, the key length adjustment factor is 0.34 and the encryption round adjustment factor is 0.36. Then the initial key length is 144 multiplied by 0.34 and rounded to 49 bits, and the initial encryption round is 14 multiplied by 0.36 and rounded to 5.

[0105] The algorithm determines whether the encryption strength corresponding to the initial encryption parameters matches the security requirements of the current mine interference area type. Security requirements are defined by the encryption strength range corresponding to that interference area type: 13 to 16 for high interference areas, 10 to 12 for medium interference areas, and 4 to 8 for low interference areas. This range is set based on the security level and interference degree of different areas in the mine, using the same quantization standard as the predicted encryption strength value. Encryption strength matching is achieved through an encryption strength evaluation model, a pre-trained support vector machine model with a radial basis function kernel (RBF) parameter of 0.65, a penalty coefficient of 1.6, and a slack variable of 0.12. The inputs are the initial key length and the initial encryption round number, and the output is the corresponding quantized encryption strength value. After inputting the initial key length and the initial encryption round number into the model, the model calculates the quantized encryption strength value through forward propagation. If this quantized value falls within the encryption strength range of the current interference area, a match is determined; if the quantized value is below the lower limit or above the upper limit of the range, a mismatch is determined, and an iterative adjustment process is initiated. During the iterative adjustment process, if the quantized encryption strength value is lower than the lower limit of the interval, it indicates that the encryption strength corresponding to the initial encryption parameters is insufficient, and the adjustment coefficient needs to be increased. The adjustment coefficient is updated at a fixed ratio of 1.1. If the quantized encryption strength value is higher than the upper limit of the interval, it indicates that the encryption strength is excessive, and the adjustment coefficient needs to be decreased. The adjustment coefficient is updated at a fixed ratio of 0.9. After each update of the adjustment coefficient, the new encryption parameters are re-substituted into the weighted calculation logic to obtain new encryption parameters. The new parameters are then input into the encryption strength evaluation model to calculate a new quantized encryption strength value, and it is determined whether the quantized value falls within the target interval. The maximum number of iterations is set to 10. If the matching condition is still not met after 10 iterations, the parameters of the 10th iteration are used as the initial encryption parameters to ensure that the process does not fall into an infinite loop. For example, in a high-interference region, if the quantized encryption strength value corresponding to the initial encryption parameters is 12.8, which is lower than the lower limit of the interval of 13, the key length adjustment coefficient and the encryption round number adjustment coefficient are updated to 1.1 times their original values, the new initial key length and the initial encryption round number are recalculated, and then input into the model to calculate the quantized encryption strength value until it falls within the interval of 13 to 16.

[0106] The encryption parameters that meet security requirements are determined as the final encryption parameters, including the final key length and the final number of encryption rounds. Based on these two parameters, an encryption strength adjustment scheme is generated. The scheme is stored in the form of a structured file, which includes the specific values ​​of the final encryption parameters, the corresponding interference area type, the parameter effective time, and the applicable scope. The effective time is set to within 1 second after the parameters are determined to ensure real-time response to environmental changes. The applicable scope is clearly defined as the currently defined interference area and all intrinsically safe mining equipment within that area.

[0107] This process establishes a direct mapping between the characteristics of the interference area and the encryption execution parameters by establishing a one-to-one correspondence between region-specific adjustment coefficients and encryption parameters, thus solving the problem of existing technologies lacking consideration for the spatial heterogeneity of mines. The combination of weighted calculation and iterative adjustment mechanisms ensures that encryption parameters can accurately match the security requirements of different interference areas, avoiding the problems of insufficient security in high-interference areas and resource waste in low-interference areas caused by uniform encryption configurations. Simultaneously, through explicit model parameters and calculation logic, the standardization and reproducibility of encryption strength adjustment are achieved, improving the system's adaptability and operational efficiency to the complex dynamic environment of mines.

[0108] Please see Figure 3 , Figure 3 This is a comprehensive comparison of the performance of the nonlinear mapping function before and after calibration in this application; Figure 3 The graph shows a comprehensive comparison of the performance of the nonlinear mapping function before and after calibration, comparing the predicted interference level with the actual interference level. The data distribution is presented by scatter points (blue for before calibration, red for after calibration), the ideal line, and the corresponding regression line. It can be seen that the average error decreased from 0.071 to 0.027, and the correlation coefficient increased from 0.933 to 0.991, with an overall improvement of 61.8%. This indicates that after calibration, the nonlinear mapping function significantly improved the fit between the predicted interference level and the actual interference level, and the prediction accuracy was greatly optimized.

[0109] Please see Figure 4 , Figure 4 The comparison results are as follows: (This refers to the cumulative distribution function of the prediction error of the degree of interference in this application.) Figure 4 The figure shows a comparison of the cumulative distribution function of the prediction error of the degree of disturbance before and after calibration of the nonlinear mapping function. The blue curve represents the error before calibration, and the red curve represents the error after calibration. The horizontal axis represents the prediction error of the degree of disturbance, and the vertical axis represents the cumulative probability. This illustrates that after calibration, the cumulative probability of the prediction error of the degree of disturbance increases more rapidly, meaning that the proportion of samples corresponding to small errors is higher. This reflects that the prediction error distribution after calibration is more concentrated in a smaller interval, and the prediction accuracy is effectively improved.

[0110] Please see Figure 5 The following describes a wireless command encryption system for intrinsically safe mining equipment according to an embodiment of this application. The intrinsically safe wireless command encryption system 500 for mining equipment includes:

[0111] The feature extraction module 501 is used to collect signal attenuation data and noise level data in the historical mine environment through a sensor network, obtain multi-dimensional historical environmental data, and then use the support vector machine algorithm to extract features from the multi-dimensional historical environmental data to obtain a characteristic environmental description.

[0112] The mapping construction module 502 is used to analyze the characteristics of multipath propagation paths based on the characteristic environment description, and to construct a nonlinear mapping function between the characteristics of multipath propagation paths and the degree of interference in signal transmission.

[0113] The model calibration module 503 is used to collect real-time signal attenuation data and noise level data in the current mine environment, obtain current environmental data, and perform electromagnetic interference level calibration on the nonlinear mapping function based on the noise level data in the current environmental data to obtain the calibrated mapping model.

[0114] The strength prediction module 504 is used to extract new multipath propagation path features from the current environmental data, input the new multipath propagation path features into the calibrated mapping model, and obtain the predicted value of the encryption strength level under the current environment.

[0115] The encryption configuration module 505 is used to determine whether the predicted value is lower than the preset encryption strength threshold. If so, it adjusts the key parameters of the wireless command encryption of the intrinsically safe mining equipment to generate an optimized encryption configuration.

[0116] The region classification module 506 is used to classify the dynamically changing signal attenuation data and noise level data in the mine environment based on the optimized encryption configuration and the support vector machine algorithm, to divide the interference region types and determine the region-specific adjustment coefficient corresponding to any interference region type.

[0117] The parameter adjustment module 507 is used to adjust the encryption parameters of the current intrinsically safe mining equipment wireless communication system in real time according to the specific adjustment coefficient of the region, so as to obtain an encryption strength adjustment scheme that matches the actual needs of the mine.

[0118] Through the collaborative efforts of the aforementioned components, the system constructs a full-process adaptive encryption system encompassing "environmental perception, model prediction, parameter optimization, and regional adaptation." This system achieves precise matching and real-time adjustment of wireless command encryption strength in the complex and dynamic environment of a mine.

[0119] The feature extraction module 501 collects historical and real-time environmental data through a sensor network, and generates a characteristic environmental description containing signal propagation characteristics and noise distribution through filtering, normalization, and support vector machine feature extraction, providing quantitative data support for subsequent mapping model construction. The mapping construction module 502 analyzes the multipath propagation path characteristics based on the characteristic environmental description, and constructs a nonlinear mapping relationship between path characteristics and signal interference level through a neural network model, overcoming the difficulty in quantifying the coupling effect of multipath effect and electromagnetic interference. The model calibration module 503 uses the actual electromagnetic interference corresponding to the current environmental noise level as a reference, and adjusts the mapping function weights through the minimum mean square error method to ensure that the model can keep up with the dynamic changes of the environment in real time and improve the accuracy of interference level prediction. The intensity prediction module 504 inputs the multipath propagation path characteristics extracted from the current environment into the calibrated model, and calculates the output through forward propagation. The interference level is determined and converted into a predicted encryption strength value, providing a clear demand guide for adjusting encryption parameters. The encryption configuration module 505 initiates an iterative optimization process for key parameters by comparing the predicted value with a preset threshold. It dynamically adjusts the key length and encryption rounds using a gradient descent algorithm to generate an optimized encryption configuration that meets basic security requirements. The region classification module 506 uses the optimized configuration as a benchmark to classify dynamic environmental data using a support vector machine, divides different interference region types, and calculates region-specific adjustment coefficients to achieve a refined perception of the heterogeneity of mine space. The parameter adjustment module 507 performs weighted calculation and iterative calibration of encryption parameters based on region-specific adjustment coefficients to ensure that the encryption strength accurately matches the security requirements of each region, ultimately forming a globally optimal encryption strength adjustment scheme. This solves the problem of mismatch between static encryption and dynamic environment and achieves a balance between security and resource utilization.

[0120] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A wireless command encryption method for intrinsically safe mining equipment, characterized in that, Includes the following steps: Step S101: Collect signal attenuation data and noise level data in the historical mine environment to obtain multi-dimensional historical environment data, and then use the support vector machine algorithm to extract features from the multi-dimensional historical environment data to obtain a characteristic environment description. Step S102: Extract multipath propagation path features based on the characterized environment description, and construct a nonlinear mapping function between the multipath propagation path features and the degree of interference in signal transmission; Step S103: Collect real-time signal attenuation data and noise level data in the current mine environment to obtain current environmental data. Based on the noise level data in the current environmental data, perform electromagnetic interference level calibration on the nonlinear mapping function to obtain the calibrated mapping model. Step S104: Extract new multipath propagation path features from the current environment data, input the new multipath propagation path features into the calibrated mapping model, and obtain the predicted value of the encryption strength level under the current environment. Step S105: Determine whether the predicted value is lower than the preset encryption strength threshold. If so, adjust the key parameters of the wireless command encryption of the intrinsically safe mining equipment to generate an optimized encryption configuration. Step S106: Based on the optimized encryption configuration, the support vector machine algorithm is used to classify the dynamically changing signal attenuation data and noise level data in the mine environment, divide the interference area types, and determine the area-specific adjustment coefficient corresponding to any interference area type. Step S107: Adjust the encryption parameters of the current intrinsically safe mining equipment wireless communication system in real time according to the specific adjustment coefficient of the region to obtain an encryption strength adjustment scheme that matches the actual needs of the mine.

2. The method according to claim 1, characterized in that, Step S101 includes: The signal attenuation data and noise level data in the historical mine environment are collected by a sensor network. The signal attenuation data represents the energy loss during wireless command transmission, and the noise level data represents the interference intensity of the mine environment on wireless command transmission. Filtering and normalizing signal attenuation data and noise level data from historical mining environments are performed to generate multidimensional historical environmental data. The support vector machine algorithm is used to extract features from the multidimensional historical environmental data to obtain feature vectors; A characteristic environment description is generated based on the feature vector, and the characteristic environment description includes feature information on signal propagation characteristics and noise distribution.

3. The method according to claim 1, characterized in that, Step S102 includes: Based on the characteristic environment description, multipath propagation path features are extracted to obtain the number and intensity of the multipath propagation paths; Determine whether the number of multipath propagation paths exceeds a preset path threshold. If so, use a neural network model to construct a nonlinear mapping function between the characteristics of the multipath propagation paths and the degree of interference in signal transmission.

4. The method according to claim 2, characterized in that, Step S103 includes: The current environmental data is obtained by real-time acquisition of signal attenuation data and noise level data in the current mine environment through sensor network, and then filtering and normalizing the data. Using the actual electromagnetic interference impact corresponding to the noise level data in the current environmental data as a reference, the deviation between the output value of the nonlinear mapping function and the actual electromagnetic interference impact is calculated; Based on the aforementioned deviation, the weights of the nonlinear mapping function are adjusted using the minimum mean square error method to generate a calibrated mapping model that characterizes the environmental disturbance properties.

5. The method according to claim 4, characterized in that, Step S104 includes: The support vector machine algorithm is used to extract features from the current environment data to obtain the current environment feature vector and extract new multipath propagation path features from it. The new multipath propagation path features are input into the calibrated mapping model, and the new multipath propagation path features are calculated through forward propagation to output the degree of interference to signal transmission in the current environment. The interference level is converted into a predicted value of the encryption strength level under the current environment, and the predicted value represents the security strength level required for wireless command encryption in the current mining environment.

6. The method according to claim 5, characterized in that, Step S105 includes: Set an encryption strength threshold to determine whether the current encryption strength meets the data transmission security requirements in a mining environment; Determine whether the predicted value is lower than the encryption strength threshold. If so, initiate the key parameter adjustment process for wireless command encryption of intrinsically safe mining equipment. The key parameters include key length and number of encryption rounds. Based on the difference between the predicted value and the encryption strength threshold, the adjustment direction and adjustment step size of the key parameters are determined. Based on the determined adjustment direction and adjustment step size, the key parameters are iteratively optimized using the gradient descent algorithm; Determine whether the encryption strength corresponding to the optimized key parameters exceeds the encryption strength threshold. If so, generate an optimized encryption configuration based on the optimized key parameters.

7. The method according to claim 6, characterized in that, Step S106 includes: The dynamic environmental feature vector is obtained by collecting dynamic signal attenuation data and noise level data in the mine environment in real time through sensor network and normalizing the data. Based on the baseline security reference corresponding to the optimized encryption configuration, the support vector machine algorithm is used to classify the dynamic environment feature vectors and divide them into different types of interference regions. For any type of interference area, based on the dynamic environmental feature vector corresponding to that area type and the optimized encryption configuration, the adjustment range of the wireless command encryption parameters under that area type is calculated, and the area-specific adjustment coefficient corresponding to any interference area type is determined.

8. The method according to claim 1, characterized in that, Step S107 includes: Obtain the region-specific adjustment coefficient and the encryption parameters of the current intrinsically safe mining equipment wireless communication system, wherein the encryption parameters include the key length and the number of encryption rounds; Using the region-specific adjustment coefficient as a weight, the encryption parameters are weighted and calculated to obtain preliminary encryption parameters; Determine whether the encryption strength corresponding to the initial encryption parameters matches the security requirements of the current mine interference area type. If not, iteratively adjust the encryption parameters based on the area-specific adjustment coefficient until the encryption strength meets the security requirements of the current mine interference area type. The encryption parameters that meet the security requirements are determined as the final encryption parameters, and an encryption strength adjustment scheme that matches the actual needs of the mine is generated based on the final encryption parameters.

9. A wireless command encryption system for intrinsically safe mining equipment, used to implement the wireless command encryption method for intrinsically safe mining equipment as described in any one of claims 1 to 8, characterized in that, The intrinsically safe wireless command encryption system for mining equipment includes: The feature extraction module is used to collect signal attenuation data and noise level data in the historical mine environment through a sensor network to obtain multi-dimensional historical environmental data. Then, the support vector machine algorithm is used to extract features from the multi-dimensional historical environmental data to obtain a characteristic environmental description. The mapping construction module is used to analyze the characteristics of multipath propagation paths based on the characteristic environment description, and construct a nonlinear mapping function between the multipath propagation path characteristics and the degree of interference in signal transmission; The model calibration module is used to collect real-time signal attenuation data and noise level data in the current mine environment, obtain current environmental data, and perform electromagnetic interference level calibration on the nonlinear mapping function based on the noise level data in the current environmental data to obtain the calibrated mapping model. The strength prediction module is used to extract new multipath propagation path features from the current environment data, input the new multipath propagation path features into the calibrated mapping model, and obtain a predicted value of the encryption strength level under the current environment. The encryption configuration module is used to determine whether the predicted value is lower than the preset encryption strength threshold. If so, it adjusts the key parameters of the wireless command encryption of the intrinsically safe mining equipment to generate an optimized encryption configuration. The region classification module is used to classify the dynamically changing signal attenuation data and noise level data in the mine environment based on the optimized encryption configuration and the support vector machine algorithm, to divide the interference region types and determine the region-specific adjustment coefficient corresponding to any interference region type. The parameter adjustment module is used to adjust the encryption parameters of the current intrinsically safe mining equipment wireless communication system in real time according to the specific adjustment coefficient of the region, so as to obtain an encryption strength adjustment scheme that matches the actual needs of the mine.

Images