A temperature compensation method and system based on inertial device

Abstract

The present invention relates to the field of temperature compensation technology, and in particular to a temperature compensation method and system based on inertial devices. The present invention proposes the following scheme: by calibrating the inertial device, the output characteristic data under different temperature conditions are obtained, and the compensation parameters are calculated in combination with the real-time temperature state. The method includes: modeling the output characteristic data through adaptive fuzzy reasoning, calculating the temperature dependence coefficient and the temperature nonlinear coefficient; using the grey correlation method to analyze the relationship between the filter signal and the temperature change; calculating the influencing factor and performing temperature compensation on the output signal according to the factor; using a genetic algorithm to optimize and adjust the compensation parameters, iteratively update, and finally output a temperature-compensated signal. The present invention improves the output accuracy and stability of the inertial device in a temperature-changing environment, and has strong adaptability and reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of temperature compensation, and in particular to a temperature compensation method and system based on an inertial device. Background Art

[0002] As a key sensor, inertial devices, such as accelerometers, gyroscopes, and magnetometers, are often affected by temperature changes during operation. Temperature changes can cause the output signal of inertial devices to drift, which in turn affects their accuracy and reliability. Therefore, temperature compensation technology has become an important direction for improving the performance of inertial devices.

[0003] At present, the temperature compensation technology of inertial devices mainly relies on two methods: one is to obtain the relationship between temperature and output signal through calibration experiments and perform compensation based on these data; the other is to predict and compensate the influence of temperature on output signal through mathematical models (such as linear fitting, nonlinear fitting, etc.). However, these traditional methods can usually only be used for specific working environments and temperature conditions, and the compensation effect is limited by calibration accuracy and model simplification.

[0004] For example, a Chinese patent with the authorization announcement number CN112254742B discloses an online fitting method for the temperature compensation parameters of a MEMS inertial device. The invention provides a method for online fitting of the temperature compensation parameters of a MEMS inertial device, which can be run in real time in a navigation computer, and uses the linear relationship between the gyroscope, accelerometer and temperature to fit the temperature compensation parameters for compensating the temperature drift of the inertial device. The online fitting method for temperature compensation parameters proposed in the invention is carried out in the navigation computer. Each cycle only uses the calculated data of the previous cycle, and there is no need to save a large amount of data. The navigation computer memory is not occupied, the amount of calculation is small, and the parameter fitting accuracy is consistent with the offline fitting accuracy. At the same time, a parameter verification function is added for checking the results during batch calibration to improve the calibration success rate. The number of products calibrated each time depends on the capacity of the temperature chamber, which improves the efficiency of batch production and has strong practical significance.

[0005] The above patents have the problems raised by this background technology: they usually rely on simplified mathematical models or large amounts of data training, and have problems such as poor adaptability, limited accuracy and high consumption of computing resources. To solve the above problems, this application designs a temperature compensation method and system based on inertial devices. Summary of the invention

[0006] The technical problem to be solved by the present invention is to address the deficiencies of the prior art and provide a temperature compensation method and system based on an inertial device. By calibrating the inertial device, the output characteristic data under different temperature conditions are obtained, and the compensation parameters are calculated in combination with the real-time temperature state. The method includes: modeling the output characteristic data through adaptive fuzzy reasoning, calculating the temperature dependence coefficient and the temperature nonlinear coefficient; using the grey correlation method to analyze the relationship between the filter signal and the temperature change; calculating the influencing factor and performing temperature compensation on the output signal according to the factor; using a genetic algorithm to optimize and adjust the compensation parameters, iteratively update, and finally output a temperature-compensated signal. The present invention improves the output accuracy and stability of the inertial device in a temperature-changing environment, and has strong adaptability and reliability.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A temperature compensation method based on an inertial device, the method comprising:

[0009] Calibrate the inertial device and obtain output characteristic data of the inertial device under different temperature conditions, wherein the output characteristic data includes temperature and filter signal;

[0010] By measuring the current ambient temperature, the real-time temperature status of the inertial device can be obtained;

[0011] Calculating compensation parameters based on the output characteristic data, and calculating the influence factor of the real-time temperature state on the output signal of the inertial device according to the compensation parameters;

[0012] Compensating the output signal of the inertial device according to the influencing factor to generate a temperature-compensated output signal;

[0013] The compensation parameter is adjusted according to the temperature-compensated output signal.

[0014] The calculating of compensation parameters comprises:

[0015] The variation law of output characteristic data is modeled through adaptive fuzzy reasoning, and the temperature dependence coefficient is calculated;

[0016] Taking temperature as the dependent variable of sequence change and filtering signal as the independent variable of sequence change, drawing a change curve, mapping the change curve into the grey relational space, comparing the change curve with the ideal curve, and calculating the difference sequence;

[0017] Performing nonlinear processing on the difference sequence by polynomial fitting to generate a temperature nonlinear coefficient;

[0018] According to the temperature dependence coefficient and the temperature nonlinear coefficient, the weight ratio is adjusted through the temperature change range under different temperature conditions to calculate the compensation parameter. The calculation formula of the compensation parameter is:

[0019] ,

[0020] in, represents the compensation parameter, k represents the adjustment factor, which is used to adjust the influence of temperature on the weighting coefficient, T represents the temperature variation range, represents the threshold point of temperature change, represents the temperature dependence coefficient, Represents the temperature nonlinear coefficient.

[0021] The modeling of the variation rule of the output characteristic data includes:

[0022] Establishing a fuzzy rule base, fuzzifying the temperature and filter signals, converting them into fuzzy sets, and defining the fuzzy sets through Gaussian membership functions to obtain membership functions;

[0023] According to the fuzzy rule base and the membership function, fuzzy reasoning is performed on the fuzzy set through IF-THEN rules, and the outputs of all rules are combined through weighted average method to calculate the fuzzy output;

[0024] The fuzzy output is defuzzified by a centroid method, and a temperature dependence coefficient is calculated.

[0025] The calculating, according to the compensation parameter, the influence factor of the real-time temperature state on the output signal of the inertial device comprises:

[0026] According to the real-time temperature state, the offset of the real-time temperature state relative to the reference temperature is calculated by comparing it with the reference temperature set in the calibration stage;

[0027] By quantitatively analyzing the influence of the rate and amplitude of temperature change under different temperature conditions on the output signal, the entropy value of temperature change is calculated;

[0028] According to the offset and the temperature change entropy value, the compensation parameters are coupled through weighted combination to calculate the influence factor. The calculation formula of the influence factor is:

[0029] ,

[0030] Where I represents the impact factor, Indicates the offset, represents the maximum offset, H represents the entropy value of temperature change, represents the maximum entropy value, and represents the weighting coefficient, and Indicates the temperature correction factor.

[0031] The adjusting the compensation parameter comprises:

[0032] According to the compensation parameter, an initial solution is randomly generated, and chromosome encoding is performed on the initial solution to form an initial population;

[0033] Set the number of iterations, decode the chromosomes of the initial population according to the influencing factors, and obtain the initial compensation parameters;

[0034] The fitness of each output signal in the initial population is calculated according to the initial compensation parameters, the fitness is iteratively optimized until the number of iterations is reached, and the adjusted compensation parameters are output.

[0035] The calibrating of the inertial device includes:

[0036] The inertial device is fixed on the calibration platform, and each axis is parallel to the horizontal plane of the calibration platform, and the calibration platform is powered by an external power supply;

[0037] After the accelerometer, gyroscope and magnetometer of the inertial device have been outputting stably for five minutes, set the position points of the middle frame and inner frame of the calibration platform and collect data at each position point;

[0038] According to the collected data, the accelerometer, gyroscope and magnetometer of the inertial device are calibrated respectively by the data fitting method.

[0039] A temperature compensation system based on an inertial device, the system comprising a data acquisition module, a signal compensation module and a parameter adjustment module;

[0040] The data acquisition module is used to calibrate the inertial device and obtain the output characteristic data of the inertial device under different temperature conditions and the real-time temperature state of the inertial device;

[0041] The signal compensation module is provided with a temperature compensation strategy, which is used to calculate compensation parameters, calculate influence factors in combination with real-time temperature status, and perform temperature compensation on the output signal according to the influence factors;

[0042] The parameter adjustment module is used to adjust the compensation parameters according to the compensated output signal and output characteristic data.

[0043] The data acquisition module comprises:

[0044] Inertial device calibration unit, used for static and dynamic calibration of inertial devices;

[0045] The temperature acquisition unit is used to measure the current ambient temperature and transmit the real-time temperature status to the subsequent processing unit;

[0046] The inertial signal acquisition unit is used to obtain the output signals of the accelerometer, gyroscope, and magnetometer sensors from the inertial device, process the output signals and convert them into digital signals.

[0047] The signal compensation module comprises:

[0048] A compensation parameter calculation unit, used to calculate and generate corresponding compensation parameters according to a preset temperature compensation algorithm;

[0049] An influence factor calculation unit, used to calculate the influence factor of the real-time temperature state on the output signal of the inertial device;

[0050] A compensation signal generating unit, used for combining the compensation parameter with the original output signal to generate a temperature-compensated output signal;

[0051] The temperature compensation strategy includes temperature dependency logic and impact factor adjustment logic;

[0052] The temperature dependency logic is configured in the compensation parameter calculation unit, and is used to analyze and process the relationship between the output signal of the inertial device and the temperature, establish a mathematical model between the temperature and the output signal, calculate the temperature dependency coefficient, and generate corresponding compensation parameters;

[0053] The impact factor adjustment logic is configured in the impact factor calculation unit, and is used to quantitatively analyze the impact of temperature changes on the output signal by comparing the offset between the real-time temperature and the reference temperature during calibration, compare the real-time temperature state with the preset temperature change pattern, calculate the temperature change entropy value, and determine the actual impact of the rate and amplitude of temperature change on the output signal.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] The present invention combines adaptive fuzzy reasoning and grey correlation analysis methods to accurately calculate the influencing factors of temperature changes on the output signals of inertial devices, and uses genetic algorithms to optimize compensation parameters, thereby significantly improving the accuracy and real-time performance of temperature compensation, while reducing dependence on complex data processing and computing resources, and having strong adaptability and high efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0057] Figure 1 This is a flow chart of a temperature compensation method based on an inertial device according to Embodiment 1 of the present invention;

[0058] Figure 2 This is a flow chart of compensation parameter calculation in Example 1 of the present invention;

[0059] Figure 3 This is a flow chart of impact factor calculation in Example 1 of the present invention;

[0060] Figure 4 This is a module diagram of a temperature compensation system based on an inertial device according to Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0062] Embodiment 1:

[0063] See also Figure 1 The present invention provides an embodiment: a temperature compensation method based on an inertial device, wherein the specific steps of the method are as follows:

[0064] S1: Calibrate the inertial device, collect and record the output characteristic data of the inertial device under different temperature conditions;

[0065] In this step, the inertial device is placed under different temperature conditions and its output characteristic data is recorded respectively. Specifically, data collection is required at multiple temperature points to obtain the output signal characteristics related to temperature. This process involves independent testing of the accelerometer, gyroscope, magnetometer and other modules of the inertial device, and analyzing the relationship between its output and temperature changes to ensure that the impact of temperature on the output signal of the inertial device can be reflected.

[0066] S2: Obtain the current real-time temperature status of the inertial device by measuring the current ambient temperature;

[0067] In this step, a temperature sensor is used to measure the current ambient temperature and obtain the temperature information of the environment where the inertial device is located in real time. The key to this step is to ensure the measurement accuracy. The temperature sensor should be placed near the inertial device to ensure that the measured ambient temperature truly reflects the operating temperature of the inertial device.

[0068] S3: Calculate corresponding compensation parameters using the collected output characteristic data;

[0069] In this step, the output characteristic data collected in S1 is used in combination with the temperature change to calculate the compensation parameters using a specific algorithm. These compensation parameters can quantify the impact of different temperatures on the output signal of the inertial device, thus providing a basis for the subsequent signal compensation process;

[0070] S4: Calculate the influence factor of the current real-time temperature state on the output signal of the inertial device in combination with the compensation parameters;

[0071] In this step, the influence factor of the current temperature state on the output signal of the inertial device is further calculated based on the compensation parameters obtained in S3 and the real-time temperature state obtained in S2. The calculation of the influence factor may involve factors such as temperature offset, temperature change rate and amplitude, and the actual influence of temperature change on the signal is quantified through formulas or models.

[0072] S5: adjusting the output signal of the inertial device according to the calculated influence factor to generate a signal output after temperature compensation;

[0073] S6: Evaluate the temperature compensated signal and adjust the compensation parameters;

[0074] In this step, the temperature compensated output signal is compared with the ideal signal to evaluate the compensation effect. If the compensation effect is not ideal, the compensation parameters calculated by S3 are further adjusted. The compensation parameters are adjusted by multiple iterations to ensure that the temperature compensation can achieve the best effect.

[0075] The performance of inertial devices is affected by their installation angle, posture and surrounding environment. By aligning the axes of the inertial device parallel to the horizontal plane of the platform, the error caused by the change of the platform angle is minimized, which helps to ensure that the output signal of each axis correctly reflects the physical quantity it measures. The specific steps for calibrating the inertial device are as follows:

[0076] S1.1: Fix the inertial device on the calibration platform, and make each axis parallel to the horizontal plane of the calibration platform, and power the calibration platform through an external power supply;

[0077] In this step, the inertial device must first be accurately fixed on a stable calibration platform to ensure that it will not be displaced or angularly deviated during the entire calibration process. The axes of the inertial device (including the three directional axes of the accelerometer, gyroscope, and magnetometer) need to be parallel to the horizontal plane of the calibration platform to ensure the accuracy and consistency of data acquisition. The external power supply of the calibration platform is used to ensure the stable operation of the platform, especially to provide the necessary power support during motion control or adjustment. By accurately fixing the inertial device and adjusting it to a parallel state with the platform, errors caused by improper installation can be eliminated, ensuring that the output data collected subsequently has high accuracy and consistency, providing a reliable foundation for the subsequent calibration process;

[0078] S1.2: Each sensor of the inertial device (accelerometer, gyroscope and magnetometer) needs to output signals in a stable state. In order to ensure the stability and accuracy of the sensor data, the inertial device needs to run on the calibration platform for more than five minutes to achieve a stable output state. At this time, by setting the middle frame and inner frame position points on the calibration platform, the system can ensure that the inertial device is in different motion states or positions (such as different angles or different directions). Then, the system will collect data at each position point, including the output signals of each sensor's acceleration, angular velocity and magnetic field strength;

[0079] By collecting data at multiple locations, the response data of the inertial device in different directions, angles and accelerations can be fully obtained, thus ensuring the accuracy of the calibration process. During the data collection process, the system can record the sensor signals at each location in detail, providing full support for subsequent analysis and calibration;

[0080] S1.3: After collecting the signal data of each sensor of the inertial device at different positions, use the data fitting method to process and analyze the data. The data fitting method includes the use of least squares method, polynomial fitting, curve fitting and other technologies, and mathematical modeling based on the output characteristics of the sensor to obtain the correction function suitable for different temperature and environmental conditions. Each sensor will have its own fitting model, which is used to eliminate the errors caused by temperature changes and other external factors during the measurement process;

[0081] Through the data fitting method, the relationship between the output signals of each sensor of the inertial device and external factors can be accurately modeled, and an accurate calibration model can be established, so that the inertial device can maintain high measurement accuracy even in harsh environments. Temperature correction further ensures the stability and consistency of the calibration results.

[0082] See also Figure 2 , the compensation parameter calculation flow chart of the embodiment of the present invention, the specific steps of S3 are as follows:

[0083] S3.1: Model the variation law of output characteristic data through adaptive fuzzy reasoning and calculate the temperature dependence coefficient;

[0084] The output characteristic data includes the response data of the inertial device under different temperature conditions. Through fuzzy processing, the output signals under different temperatures are mapped to fuzzy sets. Adaptive fuzzy reasoning can dynamically adjust the relationship between temperature and output signal by adjusting the membership function and inference rules, thereby calculating the temperature dependence coefficient. The temperature dependence coefficient reflects the linear relationship between temperature and the output signal of the inertial device, that is, the proportion of output signal change when the temperature changes. This coefficient mainly reflects the basic influence of temperature on the signal and is an important reference coefficient in the temperature compensation process;

[0085] By calculating the temperature dependence coefficient, the linear relationship between temperature and the output signal of the inertial device can be accurately captured, thus providing a preliminary, linear correction basis for subsequent compensation. The flexibility of the fuzzy inference system is utilized to ensure that the complexity brought by temperature changes can be dealt with.

[0086] S3.2: Temperature is taken as the dependent variable and the filtered signal as the independent variable for serialization. First, the change curve between temperature and filtered signal is plotted. Then, the plotted change curve is mapped to the grey relational analysis space, and the difference between it and the ideal change curve is calculated. Grey relational analysis (GRA) can effectively evaluate and quantify the relationship between temperature and output signal. Its basic idea is to judge the similarity by comparing the correlation between different sequences.

[0087] Grey correlation analysis compares the changing rules of the actual data sequence with the ideal sequence, calculates the difference sequence, and then evaluates the correlation between the signal and temperature to obtain the grey correlation value. This method is suitable for situations where system parameters are incomplete or information is insufficient, and is an effective method for modeling complex systems;

[0088] The difference between the actual output signal and the ideal output signal is quantified by calculating the difference sequence. The difference sequence reflects the nonlinear effect of temperature change on the output signal and provides data support for subsequent nonlinear compensation.

[0089] S3.3: Use polynomial fitting to perform nonlinear processing on the difference sequence obtained in the previous step. The difference sequence reflects the nonlinear effect of temperature change on the output signal, and polynomial fitting can accurately capture this nonlinear relationship. By selecting the appropriate number and order of fitting, the generated temperature nonlinear coefficient will provide correction parameters for subsequent compensation;

[0090] The temperature nonlinear coefficient reflects the complex impact of temperature changes on the output signal, especially the unevenness of the output signal change when the temperature changes greatly. This coefficient provides the necessary nonlinear correction for subsequent compensation. By processing the difference sequence through polynomial fitting, it can effectively capture and correct the nonlinear changes caused by temperature, ensuring that high accuracy and stability compensation can be maintained in the face of complex temperature fluctuations.

[0091] S3.4: By combining the temperature dependence coefficient and the temperature nonlinear coefficient, the weight ratio under different temperature conditions is adjusted. According to the temperature variation range, the weight coefficient will be dynamically adjusted between the dependence coefficient and the nonlinear coefficient to ensure that the compensation parameters can play the best effect under different temperatures. The compensation parameters can comprehensively consider the linear and nonlinear effects of temperature, so as to achieve accurate compensation of the output signal of the inertial device. The calculation formula of the compensation parameters is:

[0092] ,

[0093] in, represents the compensation parameter, k represents the adjustment factor, which is used to adjust the influence of temperature on the weighting coefficient, T represents the temperature variation range, represents the threshold point of temperature change, represents the temperature dependence coefficient, represents the temperature nonlinear coefficient;

[0094] Specifically, the temperature dependence coefficient provides the basic linear relationship between temperature change and signal output, while the temperature nonlinear coefficient corrects the nonlinear effect caused by large temperature fluctuations. By adjusting the weight ratio between the two, the best compensation for the signal can be achieved in different temperature ranges.

[0095] The specific steps of modeling the changing rules of the output characteristic data are as follows:

[0096] S3.1.1: Establish a fuzzy rule base, perform fuzzy processing on the temperature and filter signals, convert them into fuzzy sets, and define the fuzzy sets through Gaussian membership functions to obtain membership functions;

[0097] The temperature and filter signals are converted into fuzzy sets through fuzzification processing. Fuzzification refers to converting precise data into fuzzy data, which can handle the uncertainty and ambiguity in the data. The fuzzy sets of temperature and filter signals are defined through Gaussian membership function, so that they can reflect the fuzzy characteristics of system output under different input conditions;

[0098] The Gaussian membership function is a commonly used fuzzy membership function that can describe the fuzzy characteristics of the signal in a smooth manner. Its characteristic is that for small changes in the input signal (such as temperature and filter signal), the system response is smooth and gradual, and it is suitable for occasions where the temperature has a relatively smooth effect on the output signal of the inertial device.

[0099] S3.1.2: According to the fuzzy rule base and the membership function, fuzzy reasoning is performed on the fuzzy set through IF-THEN rules, and the outputs of all rules are combined through weighted average method to calculate the fuzzy output;

[0100] The IF-THEN rule is a classic fuzzy reasoning method, and its basic form is: IF (input condition) THEN (output result). According to the rule base, the premise part (such as the fuzzy state of temperature and filter signal) and the subsequent part (derivation of temperature dependence coefficient) of each rule are combined to obtain a set of fuzzy outputs. Finally, the outputs of all rules are combined using the weighted average method to obtain the final fuzzy output;

[0101] By combining fuzzy reasoning and weighted average method with rule output, multiple solutions of fuzzy sets can be effectively integrated. This method can handle multiple possible input conditions and give the most reasonable output while maintaining adaptability to complex environmental changes.

[0102] S3.1.3: Defuzzify the fuzzy output by a centroid method and calculate the temperature dependence coefficient;

[0103] The centroid method calculates the centroid position of the output curve by integrating the fuzzy output and converts it into the actual temperature dependence coefficient. The centroid method can accurately extract the most representative temperature dependence coefficient based on the distribution of the fuzzy output. This coefficient directly affects the subsequent temperature compensation process. Through defuzzification, a clear temperature dependence coefficient can be extracted from the fuzzy output. This coefficient reflects the impact of temperature changes on the output signal of the inertial device. Defuzzification effectively avoids the system uncertainty caused by fuzzy reasoning and provides a specific compensation parameter.

[0104] See also Figure 3 , the flow chart of the impact factor calculation of the embodiment of the present invention, the specific steps of S4 are as follows:

[0105] S4.1: Compare the real-time temperature state with the reference temperature set in the calibration phase, and calculate the offset of the real-time temperature state relative to the reference temperature;

[0106] The reference temperature refers to the standard reference temperature used in the calibration process. This temperature is usually set under laboratory conditions to ensure data consistency and comparability during the calibration process of the inertial device. The offset of the real-time temperature state characterizes the difference between the temperature change in the actual working environment and the calibration conditions. This is crucial for the subsequent compensation process because temperature changes will directly affect the output signal of the inertial device. By calculating the offset, the temperature difference between the current ambient temperature and the calibration conditions can be accurately captured, thereby providing a basic parameter for temperature compensation. The calculation of the offset helps ensure that the compensation parameters can be dynamically adjusted for different temperature conditions, thereby improving the measurement accuracy of the inertial device in different working environments;

[0107] Since the sensors of inertial devices (such as accelerometers, gyroscopes, and magnetometers) are usually temperature-dependent, temperature fluctuations or changes can cause drift or errors in the output signal. Calculating the offset between the real-time temperature and the reference temperature can accurately reflect the impact of ambient temperature on device performance and provide accurate reference data for subsequent compensation algorithms.

[0108] S4.2: Calculate the entropy value of temperature change by quantitatively analyzing the influence of the rate and amplitude of temperature change under different temperature conditions on the output signal;

[0109] During the operation of inertial devices, temperature changes can cause nonlinear drift or error in the signal. The rate and amplitude of temperature change directly affect the stability of the signal output. By calculating the entropy value of temperature change, the nonlinear effect of these temperature factors on the signal can be quantitatively analyzed. The calculation of the entropy value of temperature change usually adopts time series analysis, based on the temperature change and output signal data at different time points, and the information entropy is calculated to reflect the degree of influence of temperature fluctuations on the signal;

[0110] Entropy is an important measurement tool in information theory, usually used to measure the uncertainty of a system. Here, entropy is used to measure the complexity and irregularity of the output signal caused by temperature fluctuations. The larger the entropy value, the more complex and unpredictable the impact of temperature fluctuations on the signal. The entropy value of temperature change is a measure of the complexity and uncertainty of temperature change, and usually reflects the degree of influence of temperature change on signal output. By calculating the entropy value, we can quantitatively understand the amplitude and speed of signal fluctuations under different temperature change conditions, and further analyze the nonlinear impact of temperature change.

[0111] S4.3: Based on the offset and temperature change entropy values ​​calculated in the first two steps, the compensation parameters are coupled by weighted combination to calculate the impact factor. The impact factor is a comprehensive indicator that represents the comprehensive impact of the real-time temperature state and temperature change on the output signal of the inertial device. Through weighted combination, the weights of the compensation parameters can be dynamically adjusted according to the temperature offset and entropy value to cope with the changes in the signal under different temperature conditions. The calculation formula of the impact factor is:

[0112] ,

[0113] Where I represents the impact factor, Indicates the offset, represents the maximum offset, H represents the entropy value of temperature change, represents the maximum entropy value, and represents the weighting coefficient, and It represents the temperature correction factor;

[0114] The design of the weighting coefficient is based on the actual situation of the impact of temperature changes on the output signal. The weighted ratio of the offset and entropy value is determined by the system requirements and experimental data. A comprehensive influencing factor is formed by weighted combination of the offset and entropy value. This factor not only takes into account the amplitude of temperature changes, but also the complexity and uncertainty of the changes. The calculation of the influencing factor comprehensively considers the amplitude and complexity of temperature changes by weighted combination of offset and entropy value. This method can capture the impact of temperature on the signal more comprehensively and provide a more accurate correction basis for the compensation algorithm. By adjusting the weight coefficient, the compensation effect can be optimized under different temperature conditions, further improving the temperature adaptability and long-term stability of the inertial device.

[0115] The specific steps of S6 are as follows:

[0116] S6.1: randomly generate an initial solution according to the compensation parameter, perform chromosome encoding on the initial solution, and form an initial population;

[0117] Based on the calculated compensation parameters, the initial solution is constructed using the genetic algorithm initialization method. These initial solutions are called chromosomes, and each chromosome represents a candidate compensation parameter combination. First, the range of the compensation parameters is set, and a certain number of initial solutions (i.e., chromosomes) are randomly generated. These initial solutions constitute the initial population;

[0118] Chromosome encoding is a method of converting each compensation parameter into a digital code. In this scheme, binary encoding is used, and each chromosome represents a different candidate parameter combination of a solution. The process of chromosome encoding is to map the compensation parameters into the encoding space (for example, binary encoding converts the numerical value into binary bits);

[0119] In the genetic algorithm, chromosomes are generated by randomly sampling the compensation parameter range to construct the initial population. Each chromosome represents a possible compensation parameter combination, and these parameter combinations are optimized through subsequent genetic operations such as selection, crossover and mutation;

[0120] Assume that there are two compensation parameters, namely temperature offset (range: -10°C to 10°C) and entropy value (range: 0.0 to 1.0). The random generation of the initial solution can include the following chromosome encoding:

[0121] Chromosome 1: temperature offset is 3.5°C, entropy value is 0.45;

[0122] Chromosome 2: The temperature offset is -2.1°C and the entropy value is 0.80. Multiple chromosomes are randomly generated to form an initial population.

[0123] S6.2: Set the number of iterations, decode the chromosomes of the initial population according to the influencing factors, and obtain the initial compensation parameters;

[0124] According to the chromosomes in the initial population, the encoding is demapped back to the actual compensation parameters through the chromosome decoding process. Chromosome decoding converts the encoded information into the value of the actual compensation parameter (for example, converting the binary code into a real value within the corresponding compensation parameter range);

[0125] In addition, after the chromosome is decoded, the fitness value is calculated based on the compensation parameters of each chromosome. The fitness value is used to evaluate the performance of the chromosome under the current temperature conditions, usually by comparing it with the error of the actual output signal. The impact factor plays a key role here. As an indicator to measure the impact of different temperature changes on the output signal, it helps to adjust the effect of the compensation parameters;

[0126] Through chromosome decoding and fitness calculation, the candidate solutions (compensation parameters) in the initial population can be converted into actual optimization targets, providing a basis for subsequent selection, crossover and mutation operations, thereby improving the accuracy and adaptability of the compensation parameters.

[0127] In genetic algorithms, fitness is a key indicator for evaluating the quality of solutions. The fitness value reflects the effect of the compensation parameter combination on the output signal compensation. As an important basis for calculating the quality of compensation parameters, the impact factor can assist in optimizing the compensation results during the decoding process and ensure the adaptability of the compensation parameters to the current temperature changes;

[0128] Assuming that the encoding of chromosome 1 is [001110], the compensation parameters obtained after decoding are temperature offset = 4°C and entropy value = 0.55. By calculating the fitness, it is found that the compensation effect caused by this parameter combination has an accuracy of 90%;

[0129] S6.3: Calculating the fitness of each output signal in the initial population according to the initial compensation parameter, iteratively optimizing the fitness until the number of iterations is reached, and outputting the adjusted compensation parameter;

[0130] In this step, the fitness values ​​are calculated by the compensation parameters decoded for each chromosome, and these fitness values ​​are optimized. In each iteration, selection, crossover and mutation operations are performed based on the fitness values ​​to optimize the current population. Specifically:

[0131] Selection operation: select individuals with better performance according to their fitness values ​​and keep them for the next generation. This process usually uses roulette selection or tournament selection.

[0132] Crossover operation: Select two chromosomes to crossover to generate a new generation of candidate solutions. The crossover operation simulates the gene recombination process in biological genetics.

[0133] Mutation operation: Make small-scale random changes to chromosomes to simulate the mutation process in biological evolution, thereby avoiding falling into the local optimal solution during the search process.

[0134] Through repeated iterations, the optimized population will gradually approach the optimal solution and eventually obtain the adjusted compensation parameters.

[0135] Through multiple generations of optimization using genetic algorithms, the optimal solution can be found from the initial compensation parameter combination. During the iterative optimization process, the genetic algorithm gradually selects the compensation parameters that best suit the current temperature change conditions by simulating the natural selection mechanism, thereby continuously improving the temperature compensation accuracy.

[0136] The goal of the genetic algorithm is to maximize the fitness value, accelerate the search process through selection and crossover operations, and avoid falling into the local optimal solution through mutation operations. The fitness value of each generation reflects the performance of the current compensation parameters. After multiple generations of iterations, the genetic algorithm can converge to the optimal compensation parameters.

[0137] Embodiment 2:

[0138] See also Figure 4 , the present invention provides an embodiment: a temperature compensation system based on an inertial device, the system comprising a data acquisition module, a signal compensation module and a parameter adjustment module;

[0139] The data acquisition module is used to calibrate the inertial device and obtain the output characteristic data of the inertial device under different temperature conditions and the real-time temperature state of the inertial device;

[0140] The signal compensation module is provided with a temperature compensation strategy, which is used to calculate compensation parameters, calculate influence factors in combination with real-time temperature status, and perform temperature compensation on the output signal according to the influence factors;

[0141] The parameter adjustment module is used to adjust the compensation parameters according to the compensated output signal and output characteristic data.

[0142] The data acquisition module comprises:

[0143] Inertial device calibration unit, used for static and dynamic calibration of inertial devices;

[0144] The temperature acquisition unit is used to measure the current ambient temperature and transmit the real-time temperature status to the subsequent processing unit;

[0145] The inertial signal acquisition unit is used to obtain the output signals of the accelerometer, gyroscope, and magnetometer sensors from the inertial device, process the output signals and convert them into digital signals.

[0146] The signal compensation module comprises:

[0147] A compensation parameter calculation unit, used to calculate and generate corresponding compensation parameters according to a preset temperature compensation algorithm;

[0148] An influence factor calculation unit, used to calculate the influence factor of the real-time temperature state on the output signal of the inertial device;

[0149] A compensation signal generating unit, used for combining the compensation parameter with the original output signal to generate a temperature-compensated output signal;

[0150] The temperature compensation strategy includes temperature dependency logic and impact factor adjustment logic;

[0151] The temperature dependency logic is configured in the compensation parameter calculation unit, and is used to analyze and process the relationship between the output signal of the inertial device and the temperature, establish a mathematical model between the temperature and the output signal, calculate the temperature dependency coefficient, and generate corresponding compensation parameters accordingly;

[0152] The impact factor adjustment logic is configured in the impact factor calculation unit, and is used to quantitatively analyze the impact of temperature changes on the output signal by comparing the offset between the real-time temperature and the reference temperature during calibration, compare the real-time temperature state with the preset temperature change pattern, and calculate the temperature change entropy value, thereby determining the actual impact of the rate and amplitude of temperature change on the output signal.

[0153] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A temperature compensation method based on an inertial device, characterized in that: The method comprises: Calibrate the inertial device and obtain output characteristic data of the inertial device under different temperature conditions, wherein the output characteristic data includes temperature and filter signal; By measuring the current ambient temperature, the real-time temperature status of the inertial device can be obtained; Calculating compensation parameters based on the output characteristic data, and calculating the influence factor of the real-time temperature state on the output signal of the inertial device according to the compensation parameters; Compensating the output signal of the inertial device according to the influencing factor to generate a temperature-compensated output signal; adjusting the compensation parameter according to the temperature-compensated output signal; The calculating of compensation parameters comprises: The variation law of output characteristic data is modeled through adaptive fuzzy reasoning, and the temperature dependence coefficient is calculated; Taking temperature as the dependent variable of sequence change and filtering signal as the independent variable of sequence change, drawing a change curve, mapping the change curve into the grey relational space, comparing the change curve with the ideal curve, and calculating the difference sequence; Performing nonlinear processing on the difference sequence by polynomial fitting to generate a temperature nonlinear coefficient; According to the temperature dependence coefficient and the temperature nonlinear coefficient, the weight ratio is adjusted through the temperature change range under different temperature conditions to calculate the compensation parameter. The calculation formula of the compensation parameter is: ， in, represents the compensation parameter, k represents the adjustment factor, which is used to adjust the influence of temperature on the weighting coefficient, T represents the temperature variation range, represents the threshold point of temperature change, represents the temperature dependence coefficient, Represents the temperature nonlinear coefficient.

2. The temperature compensation method based on inertial device according to claim 1, characterized in that: The modeling of the variation rule of the output characteristic data includes: Establishing a fuzzy rule base, fuzzifying the temperature and filter signals, converting them into fuzzy sets, and defining the fuzzy sets through Gaussian membership functions to obtain membership functions; According to the fuzzy rule base and the membership function, fuzzy reasoning is performed on the fuzzy set through IF-THEN rules, and the outputs of all rules are combined through weighted average method to calculate the fuzzy output; The fuzzy output is defuzzified by the centroid method, and the temperature dependence coefficient is calculated.

3. The temperature compensation method based on inertial device according to claim 2, characterized in that: The calculating, according to the compensation parameter, the influence factor of the real-time temperature state on the output signal of the inertial device comprises: According to the real-time temperature state, the offset of the real-time temperature state relative to the reference temperature is calculated by comparing it with the reference temperature set in the calibration stage; By quantitatively analyzing the influence of the rate and amplitude of temperature change under different temperature conditions on the output signal, the entropy value of temperature change is calculated; According to the offset and the temperature change entropy value, the compensation parameters are coupled through weighted combination to calculate the influence factor. The calculation formula of the influence factor is: ， Where I represents the impact factor, Indicates the offset, represents the maximum offset, H represents the entropy value of temperature change, represents the maximum entropy value, and represents the weighting coefficient, and Indicates the temperature correction factor.

4. The temperature compensation method based on inertial device according to claim 1, characterized in that: The adjusting the compensation parameter comprises: According to the compensation parameter, an initial solution is randomly generated, and chromosome encoding is performed on the initial solution to form an initial population; Set the number of iterations, decode the chromosomes of the initial population according to the influencing factors, and obtain the initial compensation parameters; The fitness of each output signal in the initial population is calculated according to the initial compensation parameters, the fitness is iteratively optimized until the number of iterations is reached, and the adjusted compensation parameters are output.

5. The temperature compensation method based on inertial device according to claim 1, characterized in that: The calibrating of the inertial device includes: The inertial device is fixed on the calibration platform, and each axis is parallel to the horizontal plane of the calibration platform, and the calibration platform is powered by an external power supply; After the accelerometer, gyroscope and magnetometer of the inertial device have been outputting stably for five minutes, set the position points of the middle frame and inner frame of the calibration platform and collect data at each position point; According to the collected data, the accelerometer, gyroscope and magnetometer of the inertial device are calibrated respectively by the data fitting method.

6. A temperature compensation system based on an inertial device, used to implement a temperature compensation method based on an inertial device as claimed in any one of claims 1 to 5, characterized in that: The system includes a data acquisition module, a signal compensation module and a parameter adjustment module; The data acquisition module is used to calibrate the inertial device and obtain the output characteristic data of the inertial device under different temperature conditions and the real-time temperature state of the inertial device; The signal compensation module is provided with a temperature compensation strategy, which is used to calculate compensation parameters, calculate influence factors in combination with real-time temperature status, and perform temperature compensation on the output signal according to the influence factors; The parameter adjustment module is used to adjust the compensation parameters according to the compensated output signal and output characteristic data.

7. The temperature compensation system based on inertial devices according to claim 6, characterized in that: The data acquisition module comprises: Inertial device calibration unit, used for static and dynamic calibration of inertial devices; The temperature acquisition unit is used to measure the current ambient temperature and transmit the real-time temperature status to the subsequent processing unit; The inertial signal acquisition unit is used to obtain the output signals of the accelerometer, gyroscope, and magnetometer sensors from the inertial device, process the output signals and convert them into digital signals.

8. The temperature compensation system based on inertial devices according to claim 6, characterized in that: The signal compensation module comprises: A compensation parameter calculation unit, used to calculate and generate corresponding compensation parameters according to a preset temperature compensation algorithm; An influence factor calculation unit, used to calculate the influence factor of the real-time temperature state on the output signal of the inertial device; A compensation signal generating unit, used for combining the compensation parameter with the original output signal to generate a temperature-compensated output signal; The temperature compensation strategy includes temperature dependency logic and impact factor adjustment logic; The temperature dependency logic is configured in the compensation parameter calculation unit, and is used to analyze and process the relationship between the output signal of the inertial device and the temperature, establish a mathematical model between the temperature and the output signal, calculate the temperature dependency coefficient, and generate corresponding compensation parameters; The impact factor adjustment logic is configured in the impact factor calculation unit, and is used to quantitatively analyze the impact of temperature changes on the output signal by comparing the offset between the real-time temperature and the reference temperature during calibration, compare the real-time temperature state with the preset temperature change pattern, calculate the temperature change entropy value, and determine the actual impact of the rate and amplitude of temperature change on the output signal.

Images