Beidou receiver and edge solution method thereof

Abstract

The application relates to the technical field of geological disaster monitoring, in particular to a Beidou receiver and an edge solving method thereof, wherein the Beidou receiver comprises a Bluetooth, a main processor, a coprocessor, a Beidou module, an accelerometer, a 4G module and a FLASH; the Bluetooth is connected with the main processor and is used for device parameter configuration; the coprocessor is connected with the main processor and is used for data synchronization and main processor wake-up communication; the 4G module is connected with the coprocessor and is used for sending solving results, early warning messages and obtaining reference station data; the Beidou module and the accelerometer module are respectively connected with the coprocessor and are used for collecting data. The two processors of the main processor and the coprocessor work cooperatively, the device power consumption can be reduced, the coprocessor works under normal circumstances, the energy consumption of the device in daily operation is significantly reduced, and the power consumption is reduced by more than 30%.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geological disaster monitoring, and particularly relates to a Beidou receiver and an edge solving method thereof. BACKGROUND

[0002] Geological disaster monitoring is an important part of geological disaster prevention and engineering safety. With the successful construction of the Beidou satellite navigation system and the development of high-precision positioning technology, the Beidou system has been widely used in geological disaster monitoring such as slopes. The Beidou system can provide all-weather, all-time positioning services, and has a significant advantage in monitoring in remote areas and harsh environments. The stability of the Beidou receiver, as a key device for the application of the Beidou system, is crucial.

[0003] The Beidou receiver generally faces some challenges in field applications, especially in power supply and processing capacity. Currently, multifunctional Beidou receivers on the market usually use high-performance CPUs, which significantly increases power consumption. The overall power consumption is usually about 4W. In long periods of rainy weather, the receiver often goes offline due to power shortage, and rainy weather is also the time when geological disasters occur frequently, so the safety cannot be guaranteed. The device lacks an effective sleep mechanism, resulting in continuous power consumption even when there is no data processing demand, causing high device power consumption.

[0004] In view of the above, there is an urgent need for a Beidou receiver and an edge solving method thereof to better meet the actual application requirements. SUMMARY

[0005] The present application aims to provide a Beidou receiver and an edge solving method thereof for reducing device power consumption, and the specific technical solutions are as follows:

[0006] A Beidou receiver, comprising a Bluetooth, a main processor, a coprocessor, a Beidou module, an accelerometer, a 4G module and a FLASH;

[0007] The Bluetooth is connected to the main processor and is used for device parameter configuration;

[0008] The coprocessor is connected to the main processor and is used for data synchronization and main processor wake-up communication;

[0009] The 4G module is connected to the coprocessor and is used for sending solving results, warning messages and obtaining reference station data;

[0010] The Beidou module and the accelerometer module are respectively connected to the coprocessor and are used for data acquisition.

[0011] In addition, the present application also provides an edge solving method of a Beidou receiver for realizing data processing of the above-mentioned Beidou receiver, and the steps of the edge solving method comprise:

[0012] The Beidou receiver is turned on, and parameters of the Beidou receiver are configured; after the Beidou receiver is turned on for t minutes, the main processor, the Bluetooth and the 4G module are in a dormant state;

[0013] The coprocessor synchronously collects Beidou data and accelerometer data through the Beidou module and the accelerometer module, and stores the Beidou data and the accelerometer data in a database;

[0014] The coprocessor wakes up the 4G module at a regular time, and obtains reference station data from a cloud server through the 4G module, and stores the reference station data in the database;

[0015] Whether the Beidou receiver moves is monitored according to the accelerometer data; if the Beidou receiver moves, the main processor is woken up by the coprocessor, or a calculation time is set, and the main processor is woken up when the calculation time arrives;

[0016] After the main processor is woken up, the Beidou data and the reference station data are obtained from the database, data calculation is performed according to the Beidou data and the reference station data, and a Beidou calculation result is obtained;

[0017] The Beidou calculation result is subjected to data smoothing processing, a current smoothed result is calculated based on a smoothing weight, whether the Beidou receiver moves is judged, if the Beidou receiver does not move, the smoothing weight is determined to be in a first range, if the Beidou receiver moves, whether a Beidou result state is normal is judged based on a Beidou result state threshold, if the Beidou result state is normal, the smoothing weight is determined to be in the first range, otherwise, the smoothing weight is determined to be in a second range.

[0018] Optionally, the database is a SQLIT database in FLASH.

[0019] Optionally, while the accelerometer data is collected, an accelerometer threshold is set and whether the Beidou receiver moves is monitored based on the accelerometer threshold, and the process is as follows:

[0020] The accelerometer threshold is determined, m*60 pieces of data are obtained by dividing data of the accelerometer for m hours according to 1-minute intervals, a mean error of each piece of data is calculated, an average value of the mean errors is calculated, and n times the mean error is taken as the accelerometer threshold of each piece of data;

[0021] A difference value is obtained by subtracting the average value of the accelerometer values from the accelerometer value of 1 minute, if the difference value is greater than the accelerometer threshold, it is judged that the Beidou receiver moves, and if the difference value is less than the accelerometer threshold, the Beidou receiver is normal.

[0022] Optionally, the Beidou calculation result is subjected to data smoothing processing, and the expression is as follows:

[0023]

[0024] wherein, x i represents the i-th Beidou solution result, represents the average value of the i-k-th Beidou solution result to the i-1-th Beidou solution result, 0 < k < 24; x' i represents the current smoothed result, and θ is a smoothing weight.

[0025] Optionally, the state of the Beidou result Δ x is determined according to the following expression:

[0026]

[0027] wherein, if Δ x < σ, it is determined that the state of the Beidou result is normal, and if Δ x ≥ σ, it is determined that the state of the Beidou result is abnormal, and σ is a state threshold.

[0028] Optionally, the first range of the value of θ is 0.1 < θ < 0.5, and the second range of the value of θ is θ = 1.

[0029] Optionally, the edge solution method further comprises a pre-warning, and the process is as follows:

[0030] inputting the current smoothed result into a pre-warning model to output a model prediction value;

[0031] comparing the model prediction value with a pre-warning threshold, and if the model prediction value exceeds the pre-warning threshold, a pre-warning signal is sent out;

[0032] the main processor sends the pre-warning signal to the coprocessor, and the main processor is in a dormant state;

[0033] the coprocessor sends out the pre-warning signal, and a field sound-light alarm executes the pre-warning.

[0034] Optionally, the construction process of the pre-warning model is as follows:

[0035] data preparation, historical monitoring result data is accumulated when the Beidou receiver is running, the historical monitoring data is pre-processed, and the data format and unit are unified to obtain historical data;

[0036] an autoregressive model, a neural network model or a support vector machine model is selected according to the demand and data characteristics, an initial model is constructed, the initial model is trained by using the historical data, and a pre-warning model is obtained.

[0037] Optionally, the pre-warning model is as follows:

[0038] the autoregressive model is used in the pre-warning model, the order is determined according to the data time sequence and the parameters are estimated, the autoregressive model predicts the position change according to the historical data sequence to obtain a model prediction value;

[0039] The early warning model adopts a neural network model, is trained by a training set, is parameterized by back propagation, and is evaluated by a verification set to prevent overfitting, and the neural network is output by forward propagation to output a model prediction value.

[0040] The early warning model adopts a support vector machine model, a support vector machine selects a kernel function, maps data to find an optimal classification or regression function, and the support vector machine judges an early warning state according to a decision boundary or a regression relationship to output a model prediction.

[0041] The technical scheme of the application has the following beneficial effects:

[0042] The application integrates a high-performance main processor and a low-power coprocessor in a Beidou receiver, and the two processors work cooperatively to reduce the power consumption of the device. In a normal case, the low-power coprocessor works to significantly reduce the energy consumption of the device in daily operation, and the power consumption is reduced by more than 30%. The coprocessor is responsible for receiving data and device management, and when it is detected that no data processing is needed, the main processor can enter a sleep state to further reduce the energy consumption. In addition, when complex data processing is needed, the coprocessor can intelligently wake up the main processor to use its high-performance processing capability to complete edge calculation, reduce the dependence on a remote server, and improve the real-time performance and accuracy of data processing.

[0043] In addition to the purposes, features and advantages described above, the application has other purposes, features and advantages. The application will be further described below with reference to the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0045] Figure 1 It is a structure block diagram of the Beidou receiver in the preferred embodiment of the application;

[0046] Figure 2 It is a data processing flowchart of the main processor and the coprocessor in the preferred embodiment of the application. DETAILED DESCRIPTION

[0047] For those skilled in the technical field, the present application will be further described in detail below in combination with the drawings and specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0048] As shown in Figure 1 The embodiment provides a Beidou receiver, which comprises a Bluetooth, a main processor, a coprocessor, a Beidou module, an accelerometer, a 4G module and a FLASH.

[0049] The Bluetooth is connected with the main processor and is used for device parameter configuration.

[0050] The coprocessor is connected with the main processor and is used for data synchronization and main processor wake-up communication.

[0051] The 4G module is connected with the coprocessor and is used for sending solving results, early warning messages and obtaining reference station data.

[0052] The Beidou module and the accelerometer module are connected with the coprocessor respectively and are used for collecting data.

[0053] As shown in Figure 2 The embodiment also provides an edge solving method of a Beidou receiver, which is used for realizing data processing of the Beidou receiver.

[0054] The Beidou receiver is started, and parameter configuration is performed on the Beidou receiver; after starting for t minutes (in the embodiment, the parameter is set to be 5 minutes, and the parameter can be adjusted according to actual conditions), the main processor, the Bluetooth and the 4G module are put into sleep;

[0055] The coprocessor synchronously collects Beidou data and accelerometer data through the Beidou module and the accelerometer module, and stores the Beidou data and the accelerometer data in a database;

[0056] The coprocessor wakes up the 4G module at a set time (in the embodiment, the set time is 30 minutes, and the parameter can be adjusted according to needs, and in the embodiment, a timer is further arranged in the Beidou receiver and is used for sending a timing signal), and obtains reference station data from a cloud server through the 4G module, and stores the reference station data in the database;

[0057] Whether the Beidou receiver moves or not is monitored according to the accelerometer data, if the Beidou receiver moves, the main processor is woken up by the coprocessor, or a solving time is set, and the main processor is woken up when the solving time comes.

[0058] After the main processor wakes up, it obtains Beidou data and reference station data from the database, performs data calculation based on the Beidou data and reference station data, and obtains the Beidou calculation result;

[0059] Perform data smoothing processing on the Beidou calculation result, calculate the current smoothed result based on the smoothing weight, and determine whether the Beidou receiver has moved. If the Beidou receiver has not moved, determine that the smoothing weight is in the first range (the first range is 0.1 < θ < 0.5, and in this embodiment, θ = 0.3 is taken in the first range). If the Beidou receiver has moved, determine whether the Beidou result status is normal based on the Beidou result status threshold. If the Beidou status result shows normal, determine that the smoothing weight is in the first range, otherwise determine that the smoothing weight is in the second range (the second range preferred in this embodiment is θ = 1).

[0060] Optionally, the database is an SQLIT database in FLASH.

[0061] Optionally, while collecting accelerometer data, set the accelerometer threshold and monitor whether the Beidou receiver has moved based on the accelerometer threshold. The process is as follows:

[0062] Determine the accelerometer threshold. Divide the accelerometer data for m hours (in this embodiment, 2 hours of accelerometer data are collected, and m is an integer greater than 1, which can be determined as needed) into m * 60 segments of data at 1-minute intervals (in this embodiment, 120 segments of data). Calculate the mean error of each segment of data, which are e1, e2,..., e 120 Calculate the average value of the mean errors Use n times the mean error as the accelerometer threshold for each segment of data. In this embodiment, the accelerometer threshold is

[0063] Collect the accelerometer value a for 1 minute i , subtract the average value of the accelerometer values from the accelerometer value for 1 minute to obtain the difference If then it is determined that the Beidou receiver has moved. If then the Beidou receiver is normal.

[0064] Optionally, perform data smoothing processing on the Beidou calculation result, and the expression is as follows:

[0065]

[0066] where x i represents the i-th Beidou calculation result, represents the average value of the Beidou calculation results from the (i - k)-th to the (i - 1)-th, 0 < k < 24, and in this embodiment, the value of k is 12; x' iThe current smoothed result is represented, and θ is a smoothing weight.

[0067] Optionally, the expression for judging the state of the Beidou result is as follows:

[0068]

[0069] If Δ x <σ, the state of the Beidou result is judged to be normal, and if Δ x ≥σ, the state of the Beidou result is judged to be abnormal, and σ is a state threshold value, which is taken as σ=50mm in the embodiment, and the state threshold value can also be selected as other values according to actual needs.

[0070] Optionally, the edge solving method further comprises early warning, and the process is as follows:

[0071] The current smoothed result is input into an early warning model, and a model prediction value is output;

[0072] The model prediction value is compared with an early warning threshold value, and if the model prediction value exceeds the early warning threshold value, an early warning signal is sent out;

[0073] The main processor sends the early warning signal to the coprocessor, and the main processor is put to sleep;

[0074] The coprocessor sends out the early warning signal, and a field sound-light alarm executes early warning.

[0075] Optionally, the process of constructing the early warning model is as follows:

[0076] Data preparation: the historical monitoring result data is accumulated when the Beidou receiver is running, the historical monitoring data is preprocessed, and the data format and unit are unified to obtain historical data;

[0077] According to the requirements and data characteristics, an autoregressive model, a neural network model or a support vector machine model is selected to construct an initial model, the initial model is trained by using the historical data to obtain an early warning model.

[0078] Optionally, the early warning model is as follows:

[0079] The early warning model adopts an autoregressive model, the order is determined according to the data time sequence, and the parameters are estimated, the autoregressive model predicts the position change according to the historical data sequence to obtain a model prediction value;

[0080] The early warning model adopts a neural network model, a training set is trained, the parameters are adjusted by back propagation, and the overfitting is prevented by evaluating the validation set, and the neural network outputs a model prediction value by forward propagation;

[0081] The early warning model adopts a support vector machine model, a support vector machine selects a kernel function, maps data to find an optimal classification or regression function, the support vector machine judges an early warning state according to a decision boundary or a regression relationship, and outputs a model prediction.

[0082] The above merely describes preferred embodiments of the present application but is not intended to limit the present application. The present application can be variously changed and modified by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An edge resolution method for a BeiDou receiver, characterized in that, The edge computing method is used to implement data processing in BeiDou receivers. The steps are as follows: The Beidou receiver is powered on, and its parameters are configured. After t minutes of power-on, the main processor, Bluetooth, and 4G modules go into sleep mode. The coprocessor synchronously collects BeiDou data and accelerometer data through the BeiDou module and accelerometer module, and stores the BeiDou data and accelerometer data in the database; The coprocessor periodically wakes up the 4G module and retrieves base station data from the cloud server through the 4G module, storing the base station data in the database; The system monitors whether the BeiDou receiver has moved based on accelerometer data. If it has moved, the coprocessor wakes up the main processor. Alternatively, a calculation time is set, and the main processor is woken up when the calculation time is up. After the main processor wakes up, it retrieves BeiDou data and base station data from the database, performs data calculations based on the BeiDou data and base station data, and obtains the BeiDou calculation results. The BeiDou solution results are smoothed. The current smoothed result is calculated based on the smoothing weight. It is then determined whether the BeiDou receiver has moved. If the BeiDou receiver has not moved, the smoothing weight is set to the first range. If the BeiDou receiver has moved, the BeiDou result status is determined to be normal based on the BeiDou result status threshold. If the BeiDou status result is normal, the smoothing weight is set to the first range. Otherwise, the smoothing weight is set to the second range. While collecting accelerometer data, an accelerometer threshold is set, and the movement of the BeiDou receiver is monitored based on the accelerometer threshold. The process is as follows: To determine the accelerometer threshold, divide the accelerometer data of m hours into m*60 segments at 1-minute intervals, calculate the mean error of each segment, calculate the average mean error, and use n times the mean error as the accelerometer threshold for each segment. The difference is calculated by subtracting the average accelerometer value from the accelerometer value over one minute. If the difference is greater than the accelerometer threshold, it is determined that the BeiDou receiver has moved. If the difference is less than the accelerometer threshold, the BeiDou receiver is normal.

2. The edge resolution method for a BeiDou receiver according to claim 1, characterized in that, The database is the SQLIT database in FLASH.

3. The edge resolution method for a BeiDou receiver according to claim 1, characterized in that, The BeiDou calculation results are smoothed using the following expression: Where, x i This represents the i-th BeiDou solution result. This represents the average value from the ik-th BeiDou solution result to the (i-1)-th BeiDou solution result, where 0 <k<24;x' i This represents the current smoothed result, where θ is the smoothing weight.

4. The edge resolution method for a BeiDou receiver according to claim 3, characterized in that, Determine the status of the BeiDou results Δ x The expression is as follows: Where, if Δ x If the result is less than σ, the BeiDou result is considered normal; if it is Δ... x If the result is greater than or equal to σ, then the BeiDou result is judged to be abnormal, where σ represents the state threshold.

5. The edge resolution method for a BeiDou receiver according to claim 1, characterized in that, The first range of θ values ​​is 0.1 < θ < 0.5, and the second range of θ values ​​is θ = 1.

6. The edge resolution method for a BeiDou receiver according to claim 1, characterized in that, It also includes early warning, the process of which is as follows: Input the current smoothed result into the early warning model, and output the model's predicted value; The model's predicted value is compared with the warning threshold. If the model's predicted value exceeds the warning threshold, a warning signal is issued. The main processor sends a warning signal to the coprocessor, and the main processor goes to sleep. The coprocessor issues a warning signal, and the on-site audible and visual alarms are activated.

7. The edge resolution method for a BeiDou receiver according to claim 6, characterized in that, The construction process of the early warning model is as follows: Data preparation involves accumulating historical monitoring data during BeiDou receiver operation, preprocessing the historical monitoring data, and standardizing the data format and units to obtain historical data. Based on the requirements and data characteristics, select an autoregressive model, a neural network model, or a support vector machine model, construct an initial model, train the initial model using historical data, and obtain the early warning model.

8. The edge resolution method for a BeiDou receiver according to claim 7, characterized in that, The early warning model is as follows: The early warning model adopts an autoregressive model, which determines the order and estimates the parameters based on the data time series. The autoregressive model predicts position changes based on historical data series and obtains the model prediction value. The early warning model adopts a neural network model, which is trained on a training set, tuned through backpropagation, and evaluated on a validation set to prevent overfitting. The neural network outputs the model's predicted value through forward propagation. The early warning model adopts a support vector machine model. The support vector machine selects a kernel function, maps the data to find the optimal classification or regression function, and determines the early warning status based on the decision boundary or regression relationship, and outputs the model prediction.

Images