Multi-sensor fusion tool state prediction method

Abstract

The invention discloses a tool state prediction method based on multi-sensor fusion, and the method comprises the steps: 1, determining a monitoring quantity and a state quantity sensed by a tool, and collecting a data set of the monitoring quantity; step 2, respectively constructing a conversion equation of each monitoring quantity; and step 3, based on the conversion equation of each monitoring quantity, constructing a multi-sensor fusion tool state prediction model based on Kalman filtering so as to obtain the prediction quantity of the tool state through the tool state prediction model. According to the method, the problem that the explanatability and accuracy of a simulation result are limited due to the fact that a tool design deformation error is usually obtained according to simulation calculation in an existing tool deformation monitoring mode is solved.

Description

Technical Field

[0001] This invention relates to the field of process equipment condition monitoring technology in aircraft assembly, and more particularly to a multi-sensor fusion method for tooling condition prediction. Background Technology

[0002] In the process of assembling aircraft wings, in order to ensure the assembly quality, jig-type tooling structures are usually used. The tooling clamps and positions the products to ensure their fixation and position information during the aircraft assembly process. However, due to the influence of different tooling materials and factory temperature changes, the tooling frame will deform to a certain extent according to the temperature changes in the factory during the actual aircraft assembly process, which will affect the positioning of the products. Especially when there are large tooling dimensions and large temperature differences in the factory, the deformation of the tooling may exceed the tolerance range of the product assembly.

[0003] However, due to the lack of monitoring of tooling temperature and deformation, tooling design deformation errors are usually obtained through simulation calculations, and the interpretability and accuracy of simulation results are limited. Summary of the Invention

[0004] The purpose of this invention is to provide a multi-sensor fusion method for predicting tooling status, in order to solve the problem that existing methods for tooling deformation monitoring usually obtain tooling design deformation errors based on simulation calculations, which have limited interpretability and accuracy of simulation results.

[0005] The technical solution of this invention is as follows: This invention provides a multi-sensor fusion method for predicting the state of tooling, characterized in that it includes: Step 1: Determine the monitoring quantities and status quantities of the tooling sensing, and collect the dataset of the monitoring quantities; Step 2: Construct the transformation equation for each monitoring quantity; Step 3: Based on the transformation equation of each monitoring quantity, construct a multi-sensor fusion tooling state prediction model based on Kalman filtering, so as to obtain the predicted quantity of tooling state through the tooling state prediction model.

[0006] Optionally, in the multi-sensor fusion tooling state prediction method described above, the monitoring quantity and state quantity sensed by the tooling in step 1 represent the measurable real-time monitoring quantity of the tooling during the aircraft assembly process and the tooling state that can be accurately estimated based on the real-time monitoring quantity, respectively. The selection of the monitored quantities and state quantities is based on the deformation principle of the tooling caused by the temperature of the factory during the assembly process. The monitored quantities sensed by the tooling include temperature and displacement, while the state quantity of the tooling is the deformation quantity characterized by temperature change, that is, the actual state of the tooling estimated based on the real-time monitored temperature and displacement. The monitored quantity datasets are represented as follows: ; ; ; in, This represents the time series of temperature data collected by the tooling. The time series of data collection for the displacement sensed by the tooling is represented; t represents the sampling time; n represents the number of samples; and V represents the observation matrix.

[0007] Optionally, in the multi-sensor fusion tooling state prediction method described above, the transformation equation for each monitored quantity constructed in step 2 is: Temperature monitoring conversion formula: ; Displacement monitoring quantity conversion formula: ; in, This represents the temperature monitoring value at time t. ΔT represents the temperature monitoring value at time t-1, and ΔT represents the temperature change from time t-1 to time t. This indicates that at time t and the temperature is Tool displacement monitoring values; Indicate z t Compared to the previous moment with temperature x t-1 Tool displacement monitoring value z t-1 Related functions.

[0008] Optionally, in the multi-sensor fusion tooling state prediction method described above, the function in the conversion equation of the displacement monitoring quantity... It is obtained based on the principle of thermal expansion and contraction of tooling materials. The calculation methods include: S1. Determine the thermal expansion coefficient a and the original size d of the tooling based on the tooling design material and size requirements. S2, Based on the principle of thermal expansion and contraction of materials, the deformation of the tooling at different temperatures is calculated as follows: ; in, This indicates that at time t and the temperature is Tool displacement monitoring values, This is the initial temperature value; S3, based on the deformation formula of the tooling at different temperatures and the conversion formula of temperature monitoring quantities, the conversion equation of the tooling state quantities is determined as follows: .

[0009] Optionally, in the multi-sensor fusion tooling state prediction method described above, the method for constructing the multi-sensor fusion tooling state prediction model based on Kalman filtering in step 3 includes: Step 31: Construct a multi-sensor fusion tooling state prediction model based on the measured dataset; Step 32: Construct the transfer equation of the covariance matrix of the tooling state prediction model to represent the uncertainty of the tooling state prediction model; Step 33: Update the tooling state prediction model through iterative calculation to obtain the predicted value of the tooling state.

[0010] Optionally, in the multi-sensor fusion tooling state prediction method described above, the construction method of the multi-sensor fusion tooling state prediction model in step 31 includes: Step 31-1, based on the measured values ​​from the multi-sensor sensors, estimate the tooling state at time t as follows: ; in, The actual tooling state at time t is the predicted state obtained through temperature monitoring and displacement monitoring values. This represents the actual temperature value at time t. This represents the actual deformation at time t; Step 31-2: Based on the conversion equation between temperature monitoring data and displacement monitoring data, construct the multi-sensor fusion tooling state prediction model as follows: ; Step 31-3, determine the state transition matrix and transition error matrix according to the tooling state prediction model: ; in, Represents the state transition matrix; This represents the transfer error matrix.

[0011] Optionally, in the multi-sensor fusion tooling state prediction method described above, the method for constructing the covariance matrix transfer equation of the tooling state prediction model in step 32 includes: Step 32-1: Calculate the covariance matrix of the two monitored quantities. The calculation formula is shown below: ; in, The covariance matrix represents the monitored quantities; This represents the covariance of the temperature monitoring data; This represents the covariance of the displacement monitoring data; This represents the covariance between temperature monitoring data and displacement monitoring data; Step 32-2, based on the covariance matrix, calculate the covariance matrix transfer equation at time t as follows: ; in, This represents the transpose of the state transition matrix; Let represent the tooling state covariance matrix at time t; This represents the tooling state covariance matrix at time t-1, with its initial values ​​being... You can match the values ​​based on experience range and measurement accuracy, or you can take a default trial value and use a diagonal matrix. The default trial value is represented as follows: .

[0012] Optionally, in the multi-sensor fusion tooling state prediction method described above, the method for updating the prediction model in step 33 includes: Step 33-1, calculate the Kalman gain of the fused sensor as follows: ; Where K represents the Kalman gain; H represents the observation matrix, with values ​​ranging from 1 to 10. R represents the monitoring noise matrix; Step 33-2, correct the tooling state prediction model. The correction method for the tooling state at time t is as follows: ; in, V(:,t) This means taking the data from the t-th column of all rows of V in the monitoring dataset, which corresponds to the temperature and displacement monitoring values ​​at time t. Step 33-3, correct the covariance matrix transfer equation. The correction method for the covariance matrix transfer equation at time t is as follows: ; in, .

[0013] Optionally, the multi-sensor fusion tooling state prediction method described above further includes: Step 33-4: By iteratively executing steps 33-1 to 33-3, the predicted value of the tooling state is obtained, which is represented as the actual tooling deformation value at different temperatures.

[0014] The beneficial effects of this invention are as follows: Addressing the deformation patterns and amounts of tooling under different temperatures, this invention proposes a multi-sensor fusion method for tooling state prediction. This method integrates sensor data on different temperatures and displacements of the tooling during assembly, utilizes the principle of thermal expansion and contraction of different materials to construct a conversion equation between displacement and temperature, and then constructs a multi-sensor fusion tooling state prediction model based on the conversion equation. Furthermore, it uses the Kalman filter principle to predict the tooling state through a state update model, thereby effectively assessing the deformation at different locations of the tooling under different temperatures. This can effectively reduce the cost of tooling integrated sensing design, promptly detect tooling deformation during assembly, and provide timely warnings. Attached Figure Description

[0015] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0016] Figure 1 A flowchart of a tooling state prediction method based on multi-sensor fusion provided by the present invention; Figure 2 The curve showing the change in monitored data; Figure 3 This is the curve representing the predicted state of the tooling. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

[0018] As explained in the background section, mid-frame tooling in aircraft assembly is mainly used for clamping and product positioning. However, due to temperature variations in the factory, the tooling deforms with temperature changes, which can affect the positioning accuracy of the product to some extent. Currently, tooling design deformation errors are usually obtained through simulation calculations, but the interpretability and accuracy of the simulation results are limited.

[0019] To effectively monitor tooling deformation during assembly, the designers of this invention considered integrating intelligent sensing sensors into the tooling design. These sensors are primarily used to monitor temperature and displacement changes at different locations within the tooling. By collecting data on temperature variations and their corresponding positional changes, the amount of tooling deformation during assembly can be effectively obtained. However, while deploying temperature and displacement sensors can monitor tooling deformation at critical locations, deploying a large number of sensors is required to assess the overall deformation of the tooling, and deformation prediction is not feasible.

[0020] To address the aforementioned issues, how to predict the tooling status based on existing measurement data, thereby effectively reducing the deployment of tooling sensors and promptly detecting tooling deformation deviations, is a prominent problem in the integrated sensing design and application of tooling.

[0021] The present invention provides the following specific embodiments, which can be combined with each other. For the same or similar concepts or processes, they may not be described again in some embodiments.

[0022] To optimize the integrated sensing design of tooling, effectively reduce the number of sensors, lower tooling design costs, and ensure product assembly quality while promptly detecting tooling deformation during assembly, this invention provides a multi-sensor fusion-based tooling state prediction method. This method, based on real-time measurement data analysis and certain prediction techniques, enables tooling deformation prediction based on temperature changes. On one hand, it provides a reference for setting tolerance ranges during tooling design; on the other hand, it effectively reduces the number of sensors in the integrated sensing design of tooling, thereby significantly lowering tooling manufacturing costs.

[0023] This invention constructs a tooling status prediction model for the aircraft assembly process, which allows for timely monitoring of tooling status and its impact on the product. Furthermore, by integrating a small number of sensors, it effectively predicts the deformation of tooling at different locations.

[0024] Based on the above analysis, embodiments of the present invention provide a multi-sensor fusion method for predicting tooling state, such as... Figure 1 As shown, it includes the following steps: Step 1: Determine the monitoring quantities and status quantities of the tooling sensing, and collect the dataset of the monitoring quantities; Step 2: Construct the transformation equation for each monitoring quantity; Step 3: Based on the transformation equation of each monitoring quantity, construct a multi-sensor fusion tooling state prediction model based on Kalman filtering, so as to obtain the predicted quantity of tooling state through the tooling state prediction model.

[0025] In one implementation of this invention, the monitoring quantity and state quantity of tooling perception in step 1 above represent the real-time monitoring quantity that can be measured during the aircraft assembly process and the tooling state that can be accurately estimated based on the real-time monitoring quantity, respectively.

[0026] It should be noted that the selection of monitoring quantities and state quantities is based on the deformation principle of the tooling caused by the influence of factory temperature during the assembly process. The monitoring quantities sensed by the tooling include temperature and displacement, while the state quantities of the tooling are the deformation quantities characterized by temperature changes, that is, the actual state of the tooling estimated based on the real-time monitored temperature and displacement. The monitoring quantity datasets are represented as follows: ; ; ; in, This represents the time series of temperature data collected by the tooling. The time series of data collection for the displacement sensed by the tooling is represented; t represents the sampling time; n represents the number of samples; and V represents the observation matrix.

[0027] In one implementation of this invention, the conversion equation for each monitoring quantity constructed in step 2 above is: Temperature monitoring conversion formula: ; Displacement monitoring quantity conversion formula: ; in, This represents the temperature monitoring value at time t. ΔT represents the temperature monitoring value at time t-1, and ΔT represents the temperature change from time t-1 to time t. This indicates that at time t and the temperature is Tool displacement monitoring values; Indicate z t Compared to the previous moment with temperature x t-1 Tool displacement monitoring value z t-1 Related functions.

[0028] In one implementation of this invention, the function in the conversion equation of the above displacement monitoring quantity... It is obtained based on the principle of thermal expansion and contraction of tooling materials. The calculation methods include: S1. Determine the thermal expansion coefficient a and the original size d of the tooling based on the tooling design material and size requirements. S2, Based on the principle of thermal expansion and contraction of materials, the deformation of the tooling at different temperatures is calculated as follows: ; in, This indicates that at time t and the temperature is Tool displacement monitoring values, This is the initial temperature value, typically 26℃, but different initial values ​​can be set according to actual measurements. S3, based on the deformation formula of the tooling at different temperatures and the conversion formula of temperature monitoring quantities, the conversion equation of the tooling state quantities is determined as follows: .

[0029] In one implementation of this invention, the method for constructing the multi-sensor fusion tooling state prediction model based on Kalman filtering in step 3 above includes: Step 31: Construct a multi-sensor fusion tooling state prediction model based on the measured dataset; Step 32: Construct the transfer equation of the covariance matrix of the tooling state prediction model to represent the uncertainty of the tooling state prediction model; Step 33: Update the tooling state prediction model through iterative calculation to obtain the predicted value of the tooling state.

[0030] In this implementation, the construction method of the multi-sensor fusion tooling state prediction model in step 31 above includes: Step 31-1, based on the measured values ​​from the multi-sensor sensors, estimate the tooling state at time t as follows: ; in, The actual tooling state at time t is the predicted state obtained through temperature monitoring and displacement monitoring values. This represents the actual temperature value at time t. This represents the actual deformation at time t; Step 31-2: Based on the conversion equation between temperature monitoring data and displacement monitoring data, construct the multi-sensor fusion tooling state prediction model as follows: ; Step 31-3, determine the state transition matrix and transition error matrix according to the tooling state prediction model: ; in, Represents the state transition matrix; This represents the transfer error matrix.

[0031] Furthermore, in this implementation method, the construction method of the covariance matrix transfer equation of the tooling state prediction model in step 32 includes: Step 32-1: Calculate the covariance matrix of the two monitored quantities. The calculation formula is shown below: ; in, The covariance matrix represents the monitored quantities; This represents the covariance of the temperature monitoring data; This represents the covariance of the displacement monitoring data; This represents the covariance between temperature monitoring data and displacement monitoring data; Step 32-2, based on the covariance matrix, calculate the covariance matrix transfer equation at time t as follows: ; in, This represents the transpose of the state transition matrix; Let represent the tooling state covariance matrix at time t; This represents the tooling state covariance matrix at time t-1, with its initial values ​​being... You can match the values ​​based on experience range and measurement accuracy, or you can take a default trial value and use a diagonal matrix. The default trial value is represented as follows: .

[0032] Furthermore, in this implementation, the methods for updating the prediction model in step 33 include: Step 33-1, calculate the Kalman gain of the fused sensor as follows: ; Where K represents the Kalman gain; H represents the observation matrix, with values ​​ranging from 1 to 10. R represents the monitoring noise matrix; Step 33-2, correct the tooling state prediction model. The correction method for the tooling state at time t is as follows: ; in, V(:,t) This means taking the data from the t-th column of all rows of V in the monitoring dataset, which corresponds to the temperature and displacement monitoring values ​​at time t. Step 33-3, correct the covariance matrix transfer equation. The correction method for the covariance matrix transfer equation at time t is as follows: ; in, .

[0033] Furthermore, it also includes: Step 33-4: By iteratively executing steps 33-1 to 33-3, the predicted value of the tooling state is obtained, which is represented as the actual tooling deformation value at different temperatures.

[0034] To address the deformation patterns and amounts of tooling fixtures under different temperatures, this invention proposes a multi-sensor fusion method for tooling state prediction. This method integrates sensor data on different temperatures and displacements of the tooling during assembly, constructs a displacement-temperature conversion equation based on the thermal expansion and contraction principle of different materials, and then builds a multi-sensor fusion tooling state prediction model based on the conversion equation. Utilizing the Kalman filter principle, the method predicts the tooling state through a state update model, thereby effectively assessing the deformation at different locations of the tooling under different temperatures. This approach can effectively reduce the cost of tooling integrated sensing design, promptly detect tooling deformation during assembly, and provide timely warnings.

[0035] The following implementation example illustrates the method for predicting tooling status using multi-sensor fusion provided by this invention.

[0036] Implementation Cases like Figure 1 As shown, the multi-sensor fusion-based tooling state prediction method provided in this implementation case includes the following steps: Step 1: Select the tooling sensing and monitoring quantities and status quantities, and collect the dataset of the monitoring quantities; where the monitoring quantities are real-time collected values, and the status quantities are predicted values; for example... Figure 2 As shown, the dataset is represented as follows: ; Step two, based on the tooling design material being steel, its coefficient of thermal expansion... Original size of monitoring point The conversion equations for temperature monitoring data and displacement monitoring data are constructed as follows: ; ; in, ; Step 3: Construction of a multi-sensor fusion tooling state prediction model based on Kalman filtering includes: Step 3-1: Based on the measured data of temperature and displacement monitoring and the transformation equation, estimate the tooling state at the current moment as follows: ; The tooling state transition matrix and error matrix for step 3-2 are as follows; ; Step 3-3 calculates the covariance matrix of the temperature monitoring values ​​and the displacement monitoring values. The calculation result is as follows: ; Step 3-4: Based on the tooling state transition matrix, calculate the covariance transfer equation. The calculation result is as follows: ; Step 3-5 calculates the Kalman gain of the fused sensor. The calculation result is as follows: ; Step 3-6: Correct the tooling state prediction model. The tooling state at the current moment is corrected as follows: ; Step 3-7: Correct the covariance matrix transfer equation. The covariance matrix at the current time step is corrected as follows: ; Step 3-8 involves iteratively executing steps 3-1 to 3-8 to obtain the predicted tooling state at different times, as shown in the following figures. Figure 3 As shown.

[0037] This implementation case uses the aircraft wing assembly process as an example. Based on the tooling monitoring temperature and displacement dataset, and utilizing the principle of material thermal expansion and contraction, a conversion equation between temperature and displacement is constructed. This conversion equation is then combined to obtain a tooling state prediction model. Kalman filtering is used to update the state prediction model, and the process is iterated continuously to achieve tooling state prediction. This algorithm is validated on 76 sets of monitoring data from the aircraft assembly process. The state prediction results are as follows: Figure 3 As shown in the figure, it can be seen that the deformation of the tooling can be predicted based on the temperature change, providing a reference for the prediction of the tooling status during the assembly process.

[0038] While the embodiments disclosed in this invention are as described above, they are merely illustrative of the embodiments to facilitate understanding of the invention and are not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and variations in the form and details of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A multi-sensor fusion tool condition prediction method, characterized by, The method comprises the following steps: Step 1, determining monitoring quantities and state quantities of tooling perception, and collecting data sets of the monitoring quantities; Step 2, respectively constructing conversion equations of each monitoring quantity; Step 3, based on the conversion equations of each monitoring quantity, constructing a tooling state prediction model based on Kalman filtering multi-sensor fusion, so as to obtain a prediction quantity of the tooling state through the tooling state prediction model.

2. The multi-sensor fusion tool condition prediction method of claim 1, wherein, The monitoring quantities and state quantities of tooling perception in the step 1 respectively represent real-time monitoring quantities of the tooling in the aircraft assembly process and tooling states which can be accurately estimated according to the real-time monitoring quantities; The selection of the monitoring quantities and state quantities is based on the deformation principle of the tooling affected by the workshop temperature in the assembly process, the monitoring quantities of tooling perception include temperature and displacement, and the state quantity of the tooling is a deformation quantity represented by temperature change, that is, the real tooling state estimated according to the real-time monitored temperature and displacement; wherein, the monitoring quantity data sets respectively represent as follows: ； ； ； wherein, represents a collection time sequence of tool perceived temperature; represents a collection time sequence of tool perceived displacement; t represents a sampling time; n represents a sampling number; V represents an observation matrix.

3. The multi-sensor fusion tool condition prediction method of claim 2, wherein, The conversion equation of each monitoring quantity constructed in the step 2 is: Temperature monitoring quantity conversion formula: ; Displacement monitoring conversion formula: ; wherein, represents a temperature monitoring value at time t, represents a temperature monitoring value at time t-1, and ΔT represents a temperature change amount from time t-1 to time t represents a tool displacement monitoring value at time t and when the temperature is x. represents z t a function related to the tool displacement monitoring value z t-1 at the previous time and when the temperature is x t-1 x.

4. The multi-sensor fusion tool condition prediction method of claim 3, wherein, The conversion equation of the displacement monitoring quantity is obtained according to the thermal expansion and cold contraction principle of the tooling material, The calculation method of the displacement monitoring quantity includes: S1, determining a tooling thermal expansion coefficient a and an original size d of the tooling according to tooling design materials and size requirements; S2, calculating a deformation quantity of the tooling at different temperatures according to the material thermal expansion and contraction principle: ； wherein, represents the tool displacement monitoring value at time t and temperature represents the tool displacement monitoring value at time t and temperature is the initial temperature value; S3, determining a conversion equation of the tooling state quantity according to the deformation quantity formula of the tooling at different temperatures and the temperature monitoring quantity conversion formula, and the conversion equation is represented as: 。 5. The multi-sensor fusion tool condition prediction method of claim 1, wherein, The construction manner of the tooling state prediction model based on Kalman filtering multi-sensor fusion in the step 3 comprises: Step 31, constructing a tooling state prediction model based on multi-sensor fusion based on the measured data sets; Step 32, constructing a covariance matrix transfer equation of the tooling state prediction model, which is used to represent the uncertainty of the tooling state prediction model; Step 33, updating the tooling state prediction model through iterative calculation to obtain a prediction quantity of the tooling state.

6. The multi-sensor fusion tool condition prediction method of claim 5, wherein, The construction manner of the tooling state prediction model based on multi-sensor fusion in the step 31 comprises: Step 31-1, estimating a tooling state at time t according to the measured values of the multi-perception sensors, and the tooling state is represented as: ； wherein, is the real tool state at time t, i.e. the predicted state obtained by the temperature monitoring value and the displacement monitoring value; denotes the real temperature value at time t, denotes the real deformation at time t; Step 31-2, constructing a tooling state prediction model based on multi-sensor fusion according to the conversion equations of the temperature monitoring quantity and the displacement monitoring quantity, and the tooling state prediction model is represented as: ； Step 31-3, determining a state transition matrix and a transition error matrix according to the tooling state prediction model, and the state transition matrix and the transition error matrix are represented as: ； wherein denotes the state transition matrix; denotes the transition error matrix.

7. The multi-sensor fusion tool condition prediction method of claim 6, wherein, The construction manner of the covariance matrix transfer equation of the tooling state prediction model in the step 32 comprises: Step 32-1, calculating the covariance matrix of the two monitoring quantities, and the calculation formula is shown as follows: ； wherein, denotes the covariance matrix of the monitored quantities; denotes the covariance of the temperature monitored quantities; denotes the covariance of the displacement monitored quantities; denotes the covariance of the temperature and displacement monitored quantities; Step 32-2, calculating a covariance matrix transfer equation at time t according to the covariance matrix, and the covariance matrix transfer equation is represented as: ； wherein, denotes the transpose of the state transition matrix; denotes the tool state covariance matrix at time t; denotes the tool state covariance matrix at time t-1, which is initialized as The diagonal matrix can be matched according to the experience range and the measurement accuracy, or the default trial value is taken, and the default trial value is represented as: 。 8. The multi-sensor fusion tool condition prediction method of claim 5, wherein, The updating manner of the prediction model in the step 33 comprises: Step 33-1, calculating a Kalman gain of the fused sensors: ； Wherein, K represents Kalman gain; H represents observation matrix, and the value is ; R represents monitoring noise matrix; Step 33-2, correcting the tooling state prediction model, and the correction manner of the tooling state at time t is represented as: ； wherein, V(:,t) represents the t-th column data of all rows of V in the monitoring data set, i.e. the temperature monitoring value and displacement monitoring value corresponding to the t-th moment. Step 33-3, correcting the covariance matrix transfer equation, and the correction manner of the covariance matrix transfer equation at time t is represented as: ； wherein .

9. The multi-sensor fusion tool condition prediction method of claim 8, wherein, Further comprising: Step 33-4, obtaining a prediction quantity of the tooling state by iteratively executing the steps 33-1 to 33-3, and the prediction quantity of the tooling state is represented as actual tooling deformation quantities at different temperatures.

Images