Remote communication method, device and equipment for mixing and clarifying tank stirring device

Abstract

The invention provides a remote communication method, device and equipment for a mixing and clarifying tank stirring device, belongs to the technical field of nuclear fuel post-processing, and solves the problem of control signal fluctuation caused by electromagnetic interference, cable aging and environment abrupt change in a complex industrial scene. The method comprises the following steps: acquiring signal quality data; preprocessing the signal quality data to obtain preprocessed signal quality data; determining predicted signal attenuation data according to the preprocessed signal quality data; comparing and analyzing the predicted signal attenuation data with a preset threshold value to obtain a control instruction; and controlling the mixing and clarifying tank to communicate according to the control instruction. According to the scheme, the anti-interference performance and the prediction precision can be improved, and the communication failure rate and the energy consumption are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of nuclear fuel post-processing, and in particular to a remote communication method, device and equipment for a mixing and settling tank stirring device. Background Art

[0002] The mixing and settling tank is a key piece of equipment in the nuclear fuel reprocessing process. The stable operation of its stirring device is directly related to the efficiency of radioactive material extraction and process safety. The tank motor, as the core power unit driving the stirring device, is usually deployed in high-radiation, high-temperature, and highly corrosive environments such as spent fuel plants. The reliability of its remote communication system faces severe challenges. Currently, the industry generally adopts a remote control solution based on the Profibus-DP protocol. This technology transmits control instructions through voltage differential signals and relies on the physical medium of twisted pair cables to build a master-slave tree-like communication network. However, actual applications have shown that this solution has significant defects in complex industrial scenarios.

[0003] First, the voltage differential signaling used by the Profibus-DP protocol is highly sensitive to electromagnetic interference. The broadband electromagnetic noise generated by the densely distributed inverters, high-voltage power supply cabinets, and radiation detection equipment within the spent fuel plant can easily couple into the communication system through the cables, causing signal waveform distortion. Experimental data shows that when cable lengths exceed 800 meters, signal attenuation can reach over 30%, and the bit error rate for control commands increases exponentially with distance. Second, the tree network topology is inherently fragile: a communication interruption at any slave node will cause all downstream devices to disconnect from the network, and redundant paths between the master and slave nodes are lacking. Furthermore, the extreme environment within the plant accelerates the aging of communication cables. For example, high temperatures embrittle the insulation layer, and gamma rays degrade the shielding layer, further exacerbating signal transmission quality. Therefore, there is an urgent need to develop a remote communication method with strong anti-interference capabilities, a robust network structure, and the ability to adapt to environmental changes. This will overcome existing technical bottlenecks and ensure the efficient and safe operation of the nuclear fuel reprocessing process. Summary of the Invention

[0004] The present invention provides a remote communication method, device and equipment for a mixing and clarifying tank stirring device, which solves the problem of control signal fluctuation caused by electromagnetic interference, cable aging and sudden environmental changes in complex industrial scenarios.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] An embodiment of the present invention provides a remote communication method for a mixing and settling tank stirring device, comprising:

[0007] Obtain signal quality data;

[0008] Preprocessing the signal quality data to obtain preprocessed signal quality data;

[0009] determining predicted signal attenuation data based on the preprocessed signal quality data;

[0010] Comparing and analyzing the predicted signal attenuation data with a preset threshold value to obtain a control instruction;

[0011] According to the control instruction, the mixing and settling tank is controlled to communicate.

[0012] Optionally, obtaining signal quality data includes:

[0013] Signal quality data is obtained through sensors arranged around the mixing and settling tank; the sensors include at least one of a signal strength sensor, an environmental sensor, and a distance sensor.

[0014] Optionally, preprocessing the signal quality data to obtain preprocessed signal quality data includes:

[0015] performing cleaning processing on the signal quality data to obtain cleaned signal quality data;

[0016] performing normalization processing on the cleaned signal quality data to obtain normalized signal quality data;

[0017] Performing time alignment processing on the normalized signal quality data to obtain preprocessed signal quality data.

[0018] Optionally, determining predicted signal attenuation data based on the preprocessed signal quality data includes:

[0019] The preprocessed signal quality data is input into a communication control model for processing to obtain predicted signal attenuation data; the communication control model filters the preprocessed signal quality data to obtain an intermediate result, and linearly transforms the intermediate result to obtain predicted signal attenuation data.

[0020] Optionally, the training and generation process of the communication control model includes:

[0021] Obtain multiple historical signal quality data;

[0022] Preprocessing the plurality of historical signal quality data to obtain a plurality of preprocessed historical signal quality data;

[0023] Randomly sampling the pre-processed historical signal quality data to obtain a training data set;

[0024] Inputting the data in the training data set into the communication control model in time sequence, wherein the hyperparameters of the communication control model include the number of input features being 5, the hidden layer dimension being 128, the number of LSTM layers being 1, the number of output features being 1, the learning rate being 0.001, the loss function being an adaptive mean absolute error loss function, and the optimizer being Nadam;

[0025] The communication control model is iteratively trained for 1000 times or the training is terminated when the error value of the iterative loss for 20 consecutive times is less than a preset threshold, thereby obtaining a trained communication control model.

[0026] Optionally, comparing and analyzing the predicted signal attenuation data with a preset threshold to obtain a control instruction includes:

[0027] When the predicted signal attenuation data is greater than a first threshold and less than a second threshold, generating a first control instruction, wherein the first control instruction is used to adjust the cutoff frequency of the filter;

[0028] When the predicted signal attenuation data is greater than or equal to a second threshold, a second control instruction is generated, where the second control instruction is used to switch the backup communication channel.

[0029] Optionally, the first control instruction is based on:

[0030] f c ′=f c ×(1+kΔ), adjust the cutoff frequency of the filter;

[0031] Among them, f c ′ is the cutoff frequency of the filter before adjustment, f c is the cutoff frequency of the adjusted filter, k is the proportional coefficient, and Δ is the predicted signal attenuation data.

[0032] An embodiment of the present invention further provides a remote communication device for a mixing and settling tank stirring device, comprising:

[0033] An acquisition module, used for acquiring signal quality data;

[0034] a processing module configured to preprocess the signal quality data to obtain preprocessed signal quality data; determine predicted signal attenuation data based on the preprocessed signal quality data; and compare and analyze the predicted signal attenuation data with a preset threshold to obtain a control instruction;

[0035] The control module is used to control the mixing and clarifying tank to communicate according to the control instruction.

[0036] An embodiment of the present invention further provides a computing device, comprising: a processor and a memory storing a computer program, wherein the computer program executes the above method when executed by the processor.

[0037] An embodiment of the present invention further provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the above method.

[0038] The technical solution of the present invention includes at least the following effects:

[0039] The above-mentioned solution of the present invention obtains signal quality data; preprocesses the signal quality data to obtain preprocessed signal quality data; determines predicted signal attenuation data based on the preprocessed signal quality data; compares and analyzes the predicted signal attenuation data with a preset threshold to obtain a control instruction; and controls the mixing and clarifying tank to communicate according to the control instruction, thereby improving anti-interference performance and prediction accuracy, and achieving a reduction in communication failure rate and energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of a remote communication method for a mixing and settling tank stirring device provided by an embodiment of the present invention;

[0041] Figure 2 This is a structural diagram of a remote communication system for a mixing and settling tank stirring device provided in an embodiment of the present invention;

[0042] Figure 3 This is a structural diagram of a remote communication device for a mixing and settling tank stirring device provided in an embodiment of the present invention;

[0043] Figure 4 It is a structural diagram of a computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0045] like Figure 1 As shown, an embodiment of the present invention provides a remote communication method for a mixing and settling tank stirring device, comprising:

[0046] Step 11, obtaining signal quality data;

[0047] Step 12: preprocessing the signal quality data to obtain preprocessed signal quality data;

[0048] Step 13, determining predicted signal attenuation data based on the preprocessed signal quality data;

[0049] Step 14: Compare and analyze the predicted signal attenuation data with a preset threshold to obtain a control instruction;

[0050] Step 15: Control the mixing and settling tank to communicate according to the control instruction.

[0051] In this example, signal quality sensors are first deployed within the communication system of the mixing and settling tank's agitator. These sensors can be installed at the signal transmitter, along the transmission path, and at the receiver. The signal transmitter serves as the front-end equipment in the remote monitoring center, while the signal receiver is the communication module in the agitator. Through multi-point data acquisition, comprehensive information about the signal status at different locations can be obtained. Signal quality parameters that require collection include signal strength, electromagnetic noise intensity, temperature, humidity, and signal transmission path length. These parameters can reflect signal quality from various perspectives, providing a basis for accurate signal status assessment.

[0052] The collected raw signals often contain outliers and noise, necessitating preprocessing to remove them. Outliers, caused by sensor failures, transient interference, and other factors, can severely impact the accuracy of subsequent analysis. Noise can be addressed through filtering algorithms, such as moving average and median filtering, to effectively smooth the data and reduce the impact of random noise. Because different signal quality parameters have varying ranges and dimensions, direct analysis can result in overstating or understating the impact of certain parameters. Therefore, the cleaned data must be normalized to map the values of each parameter to a uniform interval.

[0053] After obtaining the preprocessed signal quality data, it is necessary to select a suitable prediction model to determine the signal attenuation data. For example, if the signal attenuation is related to multiple factors such as the operating state of the stirring device and the ambient temperature, and there is a complex nonlinear relationship between these factors, a neural network model can be selected for processing. The neural network model needs to be trained using the historical preprocessed signal quality data. During the training process, the data set is divided into a training set and a test set, usually divided according to a certain ratio (such as 7:3 or 8:2). The training set is used to learn and adjust the model parameters, and the test set is used to evaluate the prediction performance of the model. By continuously adjusting the parameters of the model, the prediction error of the model on the training set is minimized, and the generalization ability of the model is verified on the test set. The preprocessed signal quality data at the current moment is input into the trained prediction model to obtain the predicted signal attenuation data for a period of time in the future. The predicted time span can be set according to actual needs, such as predicting the signal attenuation in the next 5 seconds, 10 seconds or 30 seconds.

[0054] After obtaining the predicted signal attenuation data, it is compared and analyzed with the preset threshold. Among them, the preset threshold is determined based on the performance requirements and actual operating experience of the remote communication system of the mixing and settling tank stirring device; generally, the preset threshold can be determined through experimental testing, expert experience or historical data analysis, and dynamically adjusted according to actual conditions. During the comparative analysis, if the predicted signal attenuation data does not exceed the preset threshold, it means that the current communication signal quality is good and can meet the remote communication requirements of the mixing and settling tank stirring device. At this time, no special control measures are required; if the predicted signal attenuation data exceeds the preset threshold, it indicates that there is a problem with the communication signal quality, and it is necessary to generate corresponding control instructions based on the results of the comparative analysis. Control instructions can be divided into: communication parameter adjustment instructions and communication link switching instructions.

[0055] The generated control command is transmitted to the control system of the mixer-settler via a communication link. Upon receiving the control command, the mixer-settler control system parses it, identifying its type and specific content. Depending on the command type, it then takes appropriate control actions. For example, if it's a communication parameter adjustment command, the control system adjusts the relevant parameter settings of the communication module. If it's a communication link switch command, the control system initiates the link switching process, switching the communication connection from the current link to the target link.

[0056] The structure of remote communication system of mixing and clarifying tank stirring device is as follows Figure 2As shown, the controller's signal input is electrically connected to the signal outputs of multiple control consoles, the controller's signal output is electrically connected to the signal inputs of multiple frequency converters, and the frequency converter's signal output is electrically connected to the mixer-settler motor. Each frequency converter is electrically connected to only one mixer-settler motor. Specifically, the control console serves as the interface between the operator and the system and is typically equipped with operating elements such as a display, buttons, and knobs. The operator uses the console to input control commands, such as setting parameters like agitation speed and run time, and can also view real-time operating status information about the mixer-settler, such as motor speed, agitation duration, and operating mode. The controller handles the core information processing and command forwarding tasks, receiving signal inputs from multiple control consoles, analyzing, evaluating, and integrating these signals, and generating corresponding control commands based on pre-set control logic. These commands are then accurately transmitted to the corresponding frequency converter. The frequency converter is a key device connecting the controller and the motor, performing signal conversion and motor speed regulation. It receives signals from the controller and converts them into voltage and frequency signals suitable for driving the motor, enabling precise speed regulation of the mixer-settler motor. Each VFD is connected one-to-one with a mixer-settler motor. This dedicated connection ensures stable and accurate signal transmission. It independently and precisely controls the corresponding motor according to control commands, meeting the diverse speed requirements of different mixing processes. The motor is the power source of the agitator, driving the agitator blades to thoroughly mix and clarify the materials in the mixer-settler. Under the control of the VFD, the motor can run stably at the set speed, providing reliable power support for the mixing process and ensuring that the mixer-settler can complete its production tasks efficiently and stably.

[0057] The above embodiments of the present invention have the following technical effects:

[0058] (1) Ensures communication stability: By acquiring and preprocessing signal quality data, the current communication signal status can be accurately grasped, noise and other interference factors can be eliminated, and a reliable basis for subsequent analysis can be provided. Based on the preprocessed data, the predicted signal attenuation data can be determined, and the signal change trend can be predicted in advance to avoid communication interruption or data loss due to sudden signal attenuation, effectively ensuring the stability of remote communication of the mixing and clarification tank stirring device.

[0059] (2) Intelligent control is achieved: the predicted signal attenuation data is compared and analyzed with the preset threshold to generate control instructions. This process realizes the intelligentization of communication control. The system can automatically make decisions based on the actual signal conditions, without frequent human intervention, improving control efficiency and accuracy.

[0060] (3) Optimized communication performance: Based on the comparison results of the predicted signal attenuation data and the threshold, the mixing and clarifying tank is controlled to communicate in a targeted manner. This can avoid forced communication when the signal quality is poor, resulting in data errors or communication failures. It can also make full use of resources when the signal quality is good, optimize communication performance, and improve the timeliness and accuracy of data transmission.

[0061] (4) Improved system reliability: The entire method forms a complete closed loop from signal monitoring, analysis to control, so that the remote communication system of the mixing and clarification tank stirring device has the ability of self-perception, judgment and adjustment, and can operate stably in complex environments, thereby improving the reliability and adaptability of the system and ensuring the continuity and stability of the production process.

[0062] In an optional embodiment of the present invention, step 11 may include:

[0063] Step 111: Acquire signal quality data through sensors disposed around the mixing and settling tank; the sensors include at least one of a signal strength sensor, an environmental sensor, and a distance sensor.

[0064] In this embodiment, sensors are positioned around the mixing and settling tank's agitator. These sensors include a signal strength sensor, an environmental sensor, and a distance sensor. The signal strength sensor measures the voltage differential signal strength in real time, the environmental sensor collects interference parameters such as electromagnetic noise intensity, temperature, and humidity, and the distance sensor records the length of the signal transmission path. Remote data collection is achieved through a monitoring and data acquisition system, ensuring that the timestamps of each sensor data are aligned. The collected signal quality data is transmitted to a remote control console at a fixed frequency (e.g., once per second) to form a raw data set.

[0065] In an optional embodiment of the present invention, step 12 may include:

[0066] Step 121, cleaning the signal quality data to obtain cleaned signal quality data;

[0067] Step 122, normalizing the cleaned signal quality data to obtain normalized signal quality data;

[0068] Step 123 : performing time alignment processing on the normalized signal quality data to obtain pre-processed signal quality data.

[0069] In this embodiment, the purpose of data cleaning is to eliminate invalid and abnormal data to ensure data quality. For missing data caused by sensor failure or transmission interruption, linear interpolation is used to supplement it:

[0070]

[0071] Among them, x t For missing point data, x t-1 is the valid data before the missing point, x t+1 is the valid data after the missing point, Δt is the valid data time window, Δt t is the missing time window.

[0072] When processing abnormal data, thresholds can be set based on physical constraints, such as a signal strength range of 0-5V, and data outside the range is marked as abnormal.

[0073] To improve the efficiency of model training and data processing, the maximum-minimum normalization method can be used:

[0074]

[0075] Among them, x norm is the normalized signal quality data, x min is the minimum value of signal quality data, x max is the maximum value of the signal quality data, and x is the signal quality data after cleaning.

[0076] After obtaining the normalized signal quality data, the data is aligned according to a unified timestamp to generate a time series matrix, which is the preprocessed signal quality data; each row of the matrix contains parameters such as signal strength, distance, and noise.

[0077] In an optional embodiment proposed by the present invention, step 13 may include:

[0078] In step 131, the preprocessed signal quality data is input into a communication control model for processing to obtain predicted signal attenuation data; the communication control model filters the preprocessed signal quality data to obtain an intermediate result, and linearly transforms the intermediate result to obtain predicted signal attenuation data.

[0079] In this embodiment, the communication control model is essentially an LSTM model, a special recurrent neural network designed specifically for processing sequence data. It can capture long-term dependencies in the data and is very suitable for predicting time-varying processes such as signal attenuation. The LSTM model is composed of multiple LSTM units, each of which contains key components such as an input gate, a forget gate, an output gate, and a cell state. The input gate controls how much information from the input data at the current moment can enter the cell state; the forget gate determines how much information from the cell state at the previous moment needs to be forgotten; and the output gate controls how much information in the cell state can be output to the hidden state at the current moment. This unique gating mechanism enables the LSTM model to effectively and selectively memorize and forget information, thereby avoiding the problem of gradient vanishing or gradient exploding when processing long sequence data and better capturing long-term dependencies in the data.

[0080] When preprocessed signal quality data is input into the LSTM model, the model processes each data point sequentially in chronological order. During processing, the model dynamically updates the cell and hidden states through calculations by each gating unit based on the current input data and the hidden and cell states at the previous moment. This process is similar to "filtering" the input data. The model can automatically learn useful features and patterns in the signal quality data, ignore noise and irrelevant information, and convert the raw data into more representative and interpretable intermediate results. For example, the model may learn that within a specific time period, there is a correlation between signal strength and bit error rate. When the signal strength drops to a certain level, the bit error rate increases significantly. Through filtering, the model can highlight this key information and provide strong support for subsequent predictions.

[0081] Although the intermediate results obtained after filtering contain important features of the signal quality data, they may not directly correspond to the predicted signal attenuation data. The main purpose of linear transformation is to further transform the intermediate results and extract features, mapping them to the feature space of the predicted signal attenuation data, thereby obtaining prediction results that better meet practical needs. Linear transformation is usually implemented through a fully connected layer. Each neuron in a fully connected layer is connected to all neurons in the previous layer. The input data is linearly combined using a weight matrix and bias term, and an activation function is applied to finally output the predicted signal attenuation data.

[0082] In an optional embodiment of the present invention, the training and generation process of the communication control model includes:

[0083] Step 1311, obtaining a plurality of historical signal quality data;

[0084] Step 1312: pre-process the plurality of historical signal quality data to obtain a plurality of pre-processed historical signal quality data;

[0085] Step 1313: randomly sampling the plurality of pre-processed historical signal quality data to obtain a training data set;

[0086] Step 1314: Input the data in the training dataset into the communication control model in time sequence, wherein the hyperparameters of the communication control model include the number of input features being 5, the hidden layer dimension being 128, the number of LSTM layers being 1, the number of output features being 1, the learning rate being 0.001, the loss function being an adaptive mean absolute error loss function, and the optimizer being Nadam.

[0087] Step 1315 , the communication control model is iteratively trained for 1000 times or the training is terminated when the error value of the iterative loss for 20 consecutive times is less than a preset threshold, thereby obtaining a trained communication control model.

[0088] In this embodiment, first, the plurality of historical signal quality data are preprocessed according to the method described in step 12 to obtain the preprocessed plurality of historical signal quality data;

[0089] After obtaining multiple preprocessed historical signal quality data, treat all data as a population and assign a unique number to each data point. Then, use a random number generator to generate random numbers equal to the number of training samples required. Based on these numbers, corresponding data points are extracted from the population to form a training data set. For example, if there are 1000 preprocessed data points and 200 of them need to be extracted as training samples, 200 random integers between 1 and 1000 are generated and the data points corresponding to these integers are selected.

[0090] If the data contains subpopulations with different characteristics (such as data from different time periods or different communication environments), stratified sampling can be used to ensure that the proportion of these subpopulations in the training dataset is consistent with the overall population. First, the data is stratified according to these characteristics. Then, random sampling is performed within each stratum. Finally, the samples from each stratum are combined to form the training dataset. The data in the training dataset is then input into the communication control model in time sequence. At each time step, five signal quality-related features are received as input. These features are the elements of a row in the time series matrix, including signal strength, electromagnetic interference intensity, temperature, humidity, and transmission path length.

[0091] The model's hidden layer dimension determines the dimensions of the cell state and hidden state within the LSTM unit. A larger hidden layer dimension enables the model to learn more complex feature representations and patterns, but also increases the number of model parameters and computational complexity. A 128-dimensional hidden layer achieves a good balance between model complexity and computational efficiency, enabling the model to learn long-term dependencies in signal quality data.

[0092] Using a single-layer LSTM structure can reduce model complexity, training time, and the risk of overfitting to a certain extent. Although multi-layer LSTMs can learn more abstract feature levels, in this scenario, a single-layer LSTM, combined with appropriate hidden layer dimensions and hyperparameter settings, is able to effectively capture key information in signal quality data.

[0093] The output of the model is a predicted value, that is, the predicted signal attenuation data. For example, the signal strength attenuation at a certain point in the future can be predicted, so the number of output features is 1.

[0094] The learning rate controls the update step size of the model parameters in each iteration. A smaller learning rate (such as 0.001) can make the model more stable during training, preventing model oscillation and non-convergence caused by excessive parameter updates. Although a smaller learning rate slows down the training process, it can improve the model's convergence to the global optimal solution.

[0095] Mean Absolute Error (MAE) measures the average absolute difference between the model's predictions and the true values. Adaptive MAE improves upon traditional MAE by automatically adjusting the loss weight based on different samples or features, focusing more on samples or features that have a greater impact on the prediction, thereby improving the model's prediction accuracy.

[0096] Nadam is a variant of the Adam optimizer that combines the advantages of Nesterov momentum and the Adam algorithm. The Adam optimizer effectively handles sparse gradients and non-stationary objective functions by adaptively adjusting the learning rate of each parameter. Nesterov momentum helps the optimizer detect the direction of gradient changes before updating parameters, accelerating convergence and improving model generalization.

[0097] Because signal quality data is time series data, the model needs to input each sample in the training dataset sequentially in chronological order. At each time step, the model receives the input features of the current time step and combines them with the hidden state and cell state of the previous time step to perform calculations and update the hidden state and cell state of the current time step. This time-series input method enables the model to learn time-dependent relationships in signal quality data, such as the continuous change trend of signal strength over a period of time or the cumulative effect of bit error rate over time.

[0098] The iterative training process includes forward propagation, backpropagation and parameter update.

[0099] Forward propagation: In each iteration, the data in the training dataset is input into the model in time sequence. After calculation by the LSTM layer, the hidden state and cell state of each time step are obtained. Then, the hidden state is mapped to the output space through the fully connected layer to obtain the predicted value of the model.

[0100] Backpropagation and parameter update: Calculate the loss between the predicted value and the true value (using the AMAE loss function), then pass the loss gradient from the output layer to the input layer through the backpropagation algorithm to calculate the gradient of each parameter. Finally, use the Nadam optimizer to update the model parameters based on the gradient information to minimize the loss function.

[0101] There are two conditions for the end of training:

[0102] (1) Reaching a fixed number of iterations, that is, setting the model to train for 1000 iterations. During the training process, as the number of iterations increases, the model parameters are continuously adjusted, and the loss function value usually gradually decreases. When 1000 iterations are reached, training is stopped regardless of whether the loss function value has converged to a small value. This method is suitable for scenarios with clear requirements for training time or preliminary expectations for model performance.

[0103] (2) Meeting the loss error threshold: Training ends when the loss error value of 20 consecutive iterations is less than the preset threshold. The preset threshold is a smaller value set based on actual needs and model performance requirements. When the loss function value of the model in 20 consecutive iterations is less than this threshold, it means that the model parameters have basically converged, and further training will have very limited impact on the improvement of model performance. At this time, training can be stopped to obtain the trained communication control model. This method can ensure that the model stops training after achieving good performance, avoiding unnecessary iterations.

[0104] In an optional embodiment of the present invention, step 14 may include:

[0105] Step 141: When the predicted signal attenuation data is greater than a first threshold and less than a second threshold, generate a first control instruction, wherein the first control instruction is used to adjust the cutoff frequency of the filter;

[0106] Step 142: When the predicted signal attenuation data is greater than or equal to a second threshold, generate a second control instruction, where the second control instruction is used to switch to a backup communication channel.

[0107] In this embodiment, when the predicted signal attenuation data exceeds the first threshold, it indicates that the signal quality has begun to decline to a certain extent. However, at this time, communication can still maintain basically normal, except that there may be some slight interference or signal fluctuations. If no measures are taken, the signal attenuation may be further exacerbated. At this time, the cutoff frequency of the filter can be adjusted. By adjusting the cutoff frequency of the filter, the signal frequency range allowed to pass by the filter can be changed, filtering out more interference signals while retaining more valid communication signals, thereby improving the signal-to-noise ratio of the signal and improving communication quality. The first control instruction is based on:

[0108] f c ′=f c ×(1+kΔ), adjust the cutoff frequency of the filter;

[0109] Among them, f c ′ is the cutoff frequency of the filter before adjustment, f c is the cutoff frequency of the adjusted filter, k is the proportional coefficient, and Δ is the predicted signal attenuation data.

[0110] When the predicted signal attenuation data is greater than or equal to the second threshold, it indicates that the signal quality of the current communication channel can no longer meet the needs of normal communication. At this point, simply adjusting the cutoff frequency of the filter is no longer enough to solve the problem, because the signal attenuation may have reached the limit that the communication protocol or equipment can withstand. Continuing to use this channel may cause a complete interruption of communication, making it impossible to guarantee remote control of the stirring device. Therefore, more thorough measures must be taken, namely switching to a backup communication channel: first, detect the availability of the backup communication channel, including checking the channel's connection status, signal strength, bit error rate and other parameters; if the backup channel is available, the system will switch the data stream originally transmitted through the main channel to the backup channel according to the preset switching strategy.

[0111] like Figure 3 As shown, the embodiment of the present invention further provides a remote communication device 30 for a mixing and settling tank stirring device, comprising:

[0112] An acquisition module 31 is used to acquire signal quality data;

[0113] The processing module 32 is configured to pre-process the signal quality data to obtain pre-processed signal quality data; determine predicted signal attenuation data based on the pre-processed signal quality data; and compare and analyze the predicted signal attenuation data with a preset threshold to obtain a control instruction;

[0114] The control module 33 is used to control the mixing and settling tank to communicate according to the control instructions.

[0115] Optionally, the acquisition module 31 is specifically configured to:

[0116] Signal quality data is obtained through sensors arranged around the mixing and settling tank; the sensors include at least one of a signal strength sensor, an environmental sensor, and a distance sensor.

[0117] Optionally, the processing module 32 is specifically configured to:

[0118] performing cleaning processing on the signal quality data to obtain cleaned signal quality data;

[0119] performing normalization processing on the cleaned signal quality data to obtain normalized signal quality data;

[0120] Performing time sequence alignment processing on the normalized signal quality data to obtain pre-processed signal quality data.

[0121] Optionally, the processing module 32 is further specifically configured to:

[0122] The preprocessed signal quality data is input into a communication control model for processing to obtain predicted signal attenuation data; the communication control model filters the preprocessed signal quality data to obtain an intermediate result, and linearly transforms the intermediate result to obtain predicted signal attenuation data.

[0123] Optionally, the training and generation process of the communication control model includes:

[0124] Obtain multiple historical signal quality data;

[0125] Preprocessing the plurality of historical signal quality data to obtain a plurality of preprocessed historical signal quality data;

[0126] Randomly sampling the pre-processed historical signal quality data to obtain a training data set;

[0127] Inputting the data in the training data set into the communication control model in time sequence, wherein the hyperparameters of the communication control model include the number of input features being 5, the hidden layer dimension being 128, the number of LSTM layers being 1, the number of output features being 1, the learning rate being 0.001, the loss function being an adaptive mean absolute error loss function, and the optimizer being Nadam;

[0128] The communication control model is iteratively trained for 1000 times or the training is terminated when the error value of the iterative loss for 20 consecutive times is less than a preset threshold, thereby obtaining a trained communication control model.

[0129] Optionally, the processing module 32 is further specifically configured to:

[0130] When the predicted signal attenuation data is greater than a first threshold and less than a second threshold, generating a first control instruction, wherein the first control instruction is used to adjust the cutoff frequency of the filter;

[0131] When the predicted signal attenuation data is greater than or equal to a second threshold, a second control instruction is generated, where the second control instruction is used to switch the backup communication channel.

[0132] Optionally, the first control instruction is based on:

[0133] f c ′=f c ×(1+kΔ), adjust the cutoff frequency of the filter;

[0134] Among them, f c ′ is the cutoff frequency of the filter before adjustment, f c is the cutoff frequency of the adjusted filter, k is the proportional coefficient, and Δ is the predicted signal attenuation data.

[0135] It should be noted that this device is a device corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0136] like Figure 4 As shown, an embodiment of the present invention further provides a computing device 40, comprising a processor 41, a memory 42, and a program or instruction stored in the memory 42 and executable on the processor 41. When executed by the processor 41, the program or instruction implements the various processes of the aforementioned embodiment of the remote communication method for a mixing and settling tank stirring device, and can achieve the same technical effects. To avoid repetition, the description thereof will not be repeated here. It should be noted that the computing device in the embodiment of the present invention includes the aforementioned mobile electronic devices and non-mobile electronic devices.

[0137] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0138] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0139] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0140] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0141] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0142] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, ROM, RAM, a magnetic disk, or an optical disk.

[0143] In addition, it should be noted that, in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it will be understood that all or any steps or components of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in hardware, firmware, software or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0144] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general-purpose device. Therefore, the purpose of the present invention can also be achieved simply by providing a program product containing program code for implementing the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but do not necessarily need to be performed in chronological order. Certain steps can be performed in parallel or independently of each other.

[0145] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A remote communication method for a mixing and settling tank stirring device, characterized in that: include: Obtain signal quality data; Preprocessing the signal quality data to obtain preprocessed signal quality data; determining predicted signal attenuation data based on the preprocessed signal quality data; Comparing and analyzing the predicted signal attenuation data with a preset threshold value to obtain a control instruction; According to the control instruction, the mixing and settling tank is controlled to communicate.

2. The remote communication method for a mixing and settling tank stirring device according to claim 1, characterized in that: The acquiring of signal quality data includes: Signal quality data is obtained through sensors arranged around the mixing and settling tank; the sensors include at least one of a signal strength sensor, an environmental sensor, and a distance sensor.

3. The remote communication method for a mixing and settling tank stirring device according to claim 1, characterized in that: Preprocessing the signal quality data to obtain preprocessed signal quality data includes: performing cleaning processing on the signal quality data to obtain cleaned signal quality data; performing normalization processing on the cleaned signal quality data to obtain normalized signal quality data; Performing time sequence alignment processing on the normalized signal quality data to obtain pre-processed signal quality data.

4. The remote communication method for a mixing and settling tank stirring device according to claim 1, characterized in that: Determining predicted signal attenuation data according to the preprocessed signal quality data includes: The preprocessed signal quality data is input into a communication control model for processing to obtain predicted signal attenuation data; the communication control model filters the preprocessed signal quality data to obtain an intermediate result, and linearly transforms the intermediate result to obtain predicted signal attenuation data.

5. The remote communication method for a mixing and settling tank stirring device according to claim 4, characterized in that: The training generation process of the communication control model includes: Obtain multiple historical signal quality data; Preprocessing the plurality of historical signal quality data to obtain a plurality of preprocessed historical signal quality data; Randomly sampling the pre-processed historical signal quality data to obtain a training data set; Inputting the data in the training data set into the communication control model in time sequence, wherein the hyperparameters of the communication control model include the number of input features being 5, the hidden layer dimension being 128, the number of LSTM layers being 1, the number of output features being 1, the learning rate being 0.001, the loss function being an adaptive mean absolute error loss function, and the optimizer being Nadam; The communication control model is iteratively trained for 1000 times or the training is terminated when the error value of the iterative loss for 20 consecutive times is less than a preset threshold, thereby obtaining a trained communication control model.

6. The remote communication method for a mixing and settling tank stirring device according to claim 1, characterized in that: Comparing and analyzing the predicted signal attenuation data with a preset threshold to obtain a control instruction includes: When the predicted signal attenuation data is greater than a first threshold and less than a second threshold, generating a first control instruction, wherein the first control instruction is used to adjust the cutoff frequency of the filter; When the predicted signal attenuation data is greater than or equal to a second threshold, a second control instruction is generated, where the second control instruction is used to switch the backup communication channel.

7. The remote communication method for a mixing and settling tank stirring device according to claim 1, characterized in that: The first control instruction is based on: f c ′=f c ×(1+kΔ), adjust the cutoff frequency of the filter; Among them, f c ′ is the cutoff frequency of the filter before adjustment, f c is the cutoff frequency of the adjusted filter, k is the proportional coefficient, and Δ is the predicted signal attenuation data.

8. A remote communication device for a mixing and settling tank stirring device, characterized in that: include: An acquisition module, used for acquiring signal quality data; a processing module, configured to preprocess the signal quality data to obtain preprocessed signal quality data; determining predicted signal attenuation data based on the preprocessed signal quality data; Comparing and analyzing the predicted signal attenuation data with a preset threshold value to obtain a control instruction; The control module is used to control the mixing and clarifying tank to communicate according to the control instruction.

9. A computing device, characterized in that include: A processor and a memory storing a computer program, wherein when the computer program is executed by the processor, the method according to any one of claims 1 to 7 is performed.

10. A computer-readable storage medium, characterized in that The device stores instructions, which, when executed on a computer, enable the computer to execute the method according to any one of claims 1 to 7.

Images