A flexible piezoelectric material resistant to thermal expansion and a method of making the same

Abstract

The application provides a flexible piezoelectric material with thermal expansion resistance and a preparation method thereof, and comprises the following steps: S1: preparing barium titanate nano powder and zirconium tungstate nano powder, ball-milling the barium titanate nano powder and the zirconium tungstate nano powder in different proportions to obtain mixed nano powder, modifying the mixed nano powder, and then combining the modified mixed nano powder with acrylic resin to obtain photosensitive composite slurry of barium titanate and zirconium tungstate in different proportions; S2: collecting the thermal expansion coefficients of the mixed nano powder of barium titanate and zirconium tungstate in different proportions obtained in step S1, using the collected data to train a random forest model, and using a differential evolution algorithm to calculate to obtain the optimal proportion of barium titanate and zirconium tungstate; and S3: obtaining the photosensitive composite slurry of barium titanate and zirconium tungstate in the optimal proportion according to the optimal proportion of barium titanate and zirconium tungstate obtained in step S2. The application has the beneficial effect of realizing the thermal expansion resistance and flexible preparation of the flexible piezoelectric material.

Description

Technical Field

[0001] This invention belongs to the field of piezoelectric materials, and in particular relates to a flexible piezoelectric material resistant to thermal expansion and its preparation method. Background Technology

[0002] Flexible piezoelectric materials are a class of materials possessing both excellent flexibility and piezoelectric properties. Their superior electromechanical conversion performance makes them suitable for applications in sensors, energy harvesting, and storage. In practical applications, flexible piezoelectric materials hold great potential for fabricating flexible sensors using photopolymerization technology. These sensors can be used in complex shapes and curved surfaces, and can withstand bending and stretching deformations without losing their piezoelectric properties. Compared to traditional rigid piezoelectric materials, flexible piezoelectric materials are more flexible and easier to integrate with other materials or components, thus offering broader application prospects.

[0003] However, in photopolymerization molding of materials, problems such as curing failure and poor adhesion between adjacent layers often occur. This leads to an inability to control subsequent layer curing and printing steps, thus affecting the dimensional accuracy of the printed parts. Through analysis of multiple sets of comparative experiments, one of the reasons is that the material undergoes thermal expansion during UV curing, causing the cured area of ​​each layer to expand and making it impossible to adhere smoothly to the printed metal substrate. This results in the final material curing failure and the inability to form a complete product. Therefore, the use of a flexible piezoelectric material with resistance to thermal expansion is of great significance in the field of photopolymerization for sensor fabrication.

[0004] Determining the optimal ratio of barium titanate to zirconium tungstate to achieve the desired coefficient of thermal expansion is a crucial issue in the preparation process. Traditional methods typically rely on experience and extensive experimentation, which is time-consuming and inefficient. Using surrogate models and optimization algorithms, the raw data on the coefficient of thermal expansion of the composite material can be analyzed. By inputting the coefficients of thermal expansion of barium titanate and zirconium tungstate in different ratios, the coefficient of thermal expansion of the composite material can be obtained. Summary of the Invention

[0005] In view of this, in order to solve the problem of thermal expansion during photopolymerization of existing flexible piezoelectric materials, which leads to printing failures of flexible sensors, this invention proposes a flexible piezoelectric material with resistance to thermal expansion and its preparation method, achieving both thermal expansion resistance and flexible fabrication of the flexible piezoelectric material.

[0006] To achieve the above objectives, the technical solution of the present invention is implemented as follows: A flexible piezoelectric material resistant to thermal expansion, comprising barium titanate, zirconium tungstate, and silane coupling agent.

[0007] The silane coupling agent is 3-(trimethoxysilyl)propyl methacrylate.

[0008] First, a method for preparing thermally resistant nanocomposite materials is proposed, comprising the following steps: Preparation of thermal expansion resistant materials: (1) Weigh out appropriate amounts of barium titanate precursor (barium carbonate BaCO3 and titanium dioxide TiO2) and zirconium tungstate precursor (tungsten trioxide WO3 and zirconium dioxide ZrO2).

[0009] (2) Place the precursor powder into a ball mill jar, add an appropriate amount of grinding media (hard alloy balls) and deionized water. Perform ball milling on the mixture to achieve thorough mixing.

[0010] (3) After ball milling, the mixture is centrifuged, dried and sieved to obtain a uniform precursor powder mixture.

[0011] (4) The mixed precursor powder was placed in a high-temperature furnace for high-temperature solid-phase reaction. Barium titanate and zirconium tungstate samples were obtained.

[0012] (5) The obtained sample was placed in a ball mill jar for ball milling again. The ball-milled nanocomposite powder was sieved to obtain a barium titanate-zirconium tungstate piezoelectric composite material resistant to thermal expansion.

[0013] Surface grafting (material modification treatment): (1) Place the prepared barium titanate-zirconium tungstate nanocomposite powder into a clean container, add anhydrous ethanol, and disperse the nanoparticles evenly.

[0014] (2) Add 3-(trimethoxysilyl)propyl methacrylate to the well dispersed nanopowder solution as a silane coupling agent to improve the surface properties of the nanocomposite material.

[0015] (3) After the grafting reaction is completed, the grafted nanoparticles are collected by centrifugation and drying.

[0016] Combined with acrylic resin: The grafted nanoparticles were mixed with acrylic resin in acetone and then vacuum-evaporated to obtain a photosensitive composite slurry with a certain degree of fluidity.

[0017] Second, after the photosensitive composite slurry is prepared, the optimal ratio of the original materials is calculated using a random forest model and optimization algorithm.

[0018] Step 1: Data Collection and Preprocessing Experimental data on barium titanate-zirconium tungstate composites were collected, including the coefficient of thermal expansion for different ratios of barium titanate and zirconium tungstate. The collected data underwent preprocessing, such as outlier removal and normalization.

[0019] Step 2: Divide the dataset into training and test sets. The collected experimental data was divided into an 80% training set and a 20% test set. The training set was used for model training, and the test set was used for model performance evaluation.

[0020] Step 3: Train the random forest model The random forest model is trained using the training set data. Model performance is optimized by adjusting the model parameters. A target accuracy is set for the model; if the accuracy target is not met, the parameters are updated and the model is retrained.

[0021] Step 4: Evaluate model performance The performance of the trained random forest model is evaluated on the test set, such as by calculating the error between the model's predicted coefficient of thermal expansion and the actual value, to determine the model's performance metrics.

[0022] Step 5: Optimize the ratio of raw materials On a trained random forest surrogate model, the optimal material ratio of barium titanate and zirconium tungstate is obtained through differential evolution algorithm, with the minimum piezoelectric performance as a constraint. The steps are as follows: (1) Initialize the population: Randomly generate a certain number of initial solutions as the initial population.

[0023] (2) Selection operation: Use roulette wheel or other selection operators to select the next generation of population.

[0024] (3) Crossover operation: Generate new solutions according to certain crossover strategies and add them to the next generation of population.

[0025] (4) Mutation operation: Use the difference vector to calculate the new solution and update the current solution to move closer to the new solution.

[0026] (5) Determine the termination condition: Determine whether the termination condition is met, such as the number of generations reaching the upper limit or the objective function value reaching the preset accuracy requirement. If the requirement is not met, continue to update the population, use the new population to replace the old population, and continue to execute step (2); if the requirement is met, output the result and obtain the optimal original material ratio.

[0027] The experimental design uses a random forest model and differential evolution algorithm to find the proportion of barium titanate-zirconium tungstate nanocomposites, so as to determine the material ratio that ensures reliable piezoelectric properties and resistance to thermal expansion.

[0028] Third, the surface-modified composite material is printed into a sensor using DLP photopolymerization technology. This includes the following steps: DLP printing prepares flexible piezoelectric composite materials with low coefficient of thermal expansion: (1) Pour the prepared photosensitive composite slurry into the liquid tank of the DLP (Digital Light Processing) 3D printer. Set the printing parameters of the DLP printer according to the designed 3D model, such as resolution, layer thickness, and photocuring time.

[0029] (2) Start the DLP printer and cure the photosensitive composite paste layer by layer until the printing of the flexible piezoelectric material is completed.

[0030] (3) Remove the printed flexible piezoelectric material from the liquid tank and perform post-processing (cleaning, removing support structures, etc.). Place the cleaned printed part into an ultraviolet curing chamber for further curing to improve the mechanical properties and dimensional stability of the material.

[0031] (4) As needed, perform post-heat treatment on the printed flexible piezoelectric material, such as annealing or sintering, to adjust its piezoelectric properties.

[0032] Compared with existing technologies, the flexible piezoelectric material resistant to thermal expansion and its preparation method described in this invention have the following advantages: 1. The flexible piezoelectric material of the present invention is based on barium titanate, a piezoelectric ceramic material. By adding zirconium tungstate, the thermal expansion resistance of the material is effectively improved, and the thermal expansion phenomenon of the flexible piezoelectric material under photocuring is avoided.

[0033] 2. Since zirconium tungstate is a material with negative thermal expansion, the combination of barium titanate and zirconium tungstate allows for the controllability of the composite material's thermal expansion coefficient. Therefore, using the thermal expansion coefficients of composite materials prepared with different ratios of barium carbonate / zirconium tungstate as raw data, a random forest surrogate model can be trained to predict the raw material ratio for a specific thermal expansion coefficient. Based on the actual application scenario of the sensor, the constraint range of the sensor's piezoelectric performance is determined, and the optimal ratio of the raw materials is obtained through a differential evolution algorithm, enabling the sensor to possess reliable piezoelectric performance and resistance to thermal expansion.

[0034] 3. Surface modification of composite piezoelectric materials, followed by bonding with flexible resins, allows for the fabrication of piezoelectric devices via methods such as DLP printing. The material's resistance to thermal expansion, flexibility, and structural tunability make it suitable for a wider range of complex applications. Photopolymerization molding methods enable precise control over the sensor's shape, offering high production efficiency and flexibility. Attached Figure Description

[0035] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the algorithm calculation in step S2 of the present invention; Figure 2 This is a schematic diagram of the differential evolution algorithm optimization in step D5 of the embodiment of the present invention. Detailed Implementation

[0036] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0037] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0038] First, barium titanate (BaTiO3) and zirconium tungstate (Zr(WO4)2) nanocomposite powders were prepared via a high-temperature solid-state reaction method. The specific experimental procedure is as follows: 1. Ingredients: First, weigh appropriate amounts of barium titanate precursor (such as barium carbonate BaCO3 and titanium dioxide TiO2) and zirconium tungstate precursor (tungsten trioxide (WO3) and zirconium dioxide (ZrO2)). The mass ratio of barium titanate to zirconium tungstate can be adjusted according to the required coefficient of thermal expansion.

[0039] 2. Mixing: Place the precursor powder into a ball mill jar, add an appropriate amount of grinding media (carbide balls) and deionized water. Ball mill the mixture to achieve thorough mixing. Set the milling time to 6 hours and the rotation speed to 300 rpm. After milling, centrifuge, dry, and sieve the mixture to obtain a uniform precursor powder.

[0040] 3. High-Temperature Solid-State Reaction: The mixed precursor powder is placed in a high-temperature furnace for a high-temperature solid-state reaction. The reaction temperature can be set to 800-1000℃, and the holding time is 4 hours. At this temperature, barium carbonate and titanium dioxide will react to form barium titanate, and sodium silicate and sodium zirconate will react to form zirconium tungstate. After the high-temperature solid-state reaction is completed, the sample is allowed to cool naturally to room temperature inside the furnace.

[0041] 4. Ball milling: The powder after the high-temperature solid-phase reaction is ball-milled again to further improve the dispersibility and uniformity of the nanoparticles. The ball milling time is set to 6 hours and the rotation speed is 300 rpm.

[0042] 5. Sieving and Storage: The ball-milled nanocomposite powder is sieved to remove any potential particle agglomeration. After sieving, the nanopowder is stored in a sealed container to prevent moisture and oxidation.

[0043] Then, the surrogate model was trained using experimental data on the thermal expansion coefficient of the composite material, and the optimal original material ratio was selected using an optimization algorithm. The specific experimental procedure is as follows: Proxy model part 1. Data collection and preprocessing Twenty sets of experimental data on barium titanate-zirconium tungstate composites were collected, including the coefficient of thermal expansion for different ratios of barium titanate and zirconium tungstate. For example, (BaTiO3 content: 30wt%, Zr(WO4)2 content: 70wt%, coefficient of thermal expansion: 0.8 ppm / °C). The collected data were preprocessed, including outlier removal and normalization.

[0044] 2. Divide the dataset into training and test sets. The 20 collected datasets were divided into two sets according to a certain ratio (e.g., 80% training set, 20% test set). The training set was used for model training, and the test set was used for model performance evaluation.

[0045] 3. Train the random forest model The random forest model is trained using the training set data. Model performance is optimized by adjusting the model parameters. A target accuracy of 5% is set for the best-case scenario. If the accuracy target is not met, the parameters are updated and the model is retrained. If the target accuracy is still not met, an ensemble model incorporating random forest, support vector machine, and extreme learning machine is built to improve prediction accuracy.

[0046] 4. Evaluate model performance The trained random forest model is evaluated on the test set, and the error between the model's predicted coefficient of thermal expansion and the actual value is calculated to determine the model's performance metrics. When the error between the predicted coefficient of thermal expansion and the actual value does not exceed 0.5, a surrogate model is output.

[0047] 5. Model Application and Validation Predicting the coefficient of thermal expansion: Given a set ratio of barium titanate and zirconium tungstate as input (e.g., BaTiO3 content: 50wt%, Zr(WO4)2 content: 50wt%), a trained random forest model was used to predict the coefficient of thermal expansion for this combination (predicted result: coefficient of thermal expansion 2.5 ppm / °C; actual data value: coefficient of thermal expansion 2.0 ppm / °C). Based on the optimal ratio of barium titanate and zirconium tungstate recommended by the random forest model, barium titanate-zirconium tungstate nanocomposite powder was prepared experimentally. The required barium titanate and zirconium tungstate precursors (e.g., barium titanate: barium carbonate BaCO3 and titanium dioxide TiO2; zirconium tungstate: zirconium tetroxide ZrO4 and ammonium tungstate (NH4)2WO4) were mixed according to the recommended ratio, and then the nanocomposite powder was prepared according to the previously mentioned experimental steps. The coefficient of thermal expansion of the obtained material was measured experimentally to verify the model's prediction accuracy. The actual measured coefficient of thermal expansion was 2.0 ppm / °C, which is close to the target value predicted by the model, indicating that the model has good prediction accuracy.

[0048] Experimental verification shows that the model predictions are quite close to the actual measurement results, providing strong theoretical guidance for the actual preparation process.

[0049] Optimization algorithm section Differential evolution algorithm to solve for the optimal ratio of original materials Based on the actual application scenarios of the sensor, a constraint range for piezoelectric performance is set, using the piezoelectric strain constant of the piezoelectric material as the indicator. A larger piezoelectric strain constant results in better piezoelectric output characteristics of the sensor, meaning a higher output voltage per unit stress. A minimum piezoelectric strain constant of 20 pC / N is set, and this is verified using a withstand voltage tester and a piezoelectric tester (static piezoelectric coefficient d). 33 The measurement instrument is used for calculation. The differential evolution algorithm is used to optimize the selection of data to obtain the original material ratio with the best thermal expansion resistance under the given piezoelectric performance requirements.

[0050] First, the minimum piezoelectric strain constant is set to 20 pC / N.

[0051] 1. Initialize the population: Generate a certain number of random solutions as the initial population. In this example, the randomly generated solution is the material ratio of barium titanate and zirconium tungstate.

[0052] 2. Selection Operation: The roulette wheel selection operator is used to select individuals with higher fitness from the current population to generate the next generation. In this example, individuals with higher fitness refer to material ratios with lower coefficients of thermal expansion.

[0053] 3. Crossover Operation: A new material ratio is calculated using a difference vector, and a new solution is generated according to a certain crossover strategy, which is then added to the next generation of the population. In this example, a mutation-crossover strategy is used: three individuals from the current population are selected, a new material ratio is obtained through difference operations, and a new solution is generated using a crossover operation.

[0054] 4. Mutation Operation: Calculate the new material mix using the difference vector and update the current solution to approximate the new solution. In this example, the difference vector can be calculated from two individuals in the current population.

[0055] 5. Determine Termination Conditions: Determine if termination conditions are met, such as reaching the upper limit of the evolutionary generation or the objective function value reaching a preset accuracy requirement. In this example, the algorithm can be terminated when the evolutionary generation reaches a certain number or the piezoelectric strain constant reaches a preset minimum value. Then, determine if the coefficient of thermal expansion meets the accuracy requirements. If not, update the population by adding the newly generated material ratio to the next generation. This replaces the old population with the new one, and step 2 continues. If the coefficient of thermal expansion requirement is met, output the result. The final original material ratio is the optimized solution, which can be used in material preparation and applications. The optimized result shows BaTiO3 content: 62wt%, Zr(WO4)2 content: 38wt%, and coefficient of thermal expansion: 4.8 ppm / °C.

[0056] When using differential evolution algorithms for material composition optimization, appropriate parameters can be set according to the specific problem. These include, for example, population size, number of generations, crossover probability, and mutation probability. Simultaneously, the objective function and constraints need to be determined and appropriately set based on the actual problem. In this example, the objective function is the coefficient of thermal expansion, and the constraint is the minimum piezoelectric performance constraint.

[0057] After determining the proportions in the composite material, the surface of the material is modified to allow it to combine with a flexible photosensitive resin, thereby enabling the DLP photocuring process to prepare the sensor.

[0058] Surface grafting: (1) Place the prepared barium titanate-zirconium tungstate nanocomposite powder into a clean container, add enough anhydrous ethanol to ensure uniform dispersion of the nanoparticles. Use an ultrasonic vibrator to disperse the powder for 30-60 minutes to ensure that the nanoparticles are fully dispersed in the ethanol. The ultrasonic vibrator parameters are 100W and 40KHz.

[0059] Concentration of nanopowders: Weigh barium carbonate (BaCO3) and titanium dioxide (TiO2) in a 1:1 molar ratio, taking 10g of barium carbonate and 5.6g of titanium dioxide (total mass 15.6g). Weigh tungsten trioxide (WO3) and zirconium dioxide (ZrO2) in a 2:1 molar ratio, taking 8.6g of tungsten oxide and 4.4g of zirconium oxide (total mass 13.6g).

[0060] (2) Add an appropriate amount of 3-(trimethoxysilyl)propyl methacrylate to the well dispersed nanopowder solution as a silane coupling agent. Mechanically stir the mixture at room temperature for 6 hours to allow the silane coupling agent to undergo a grafting reaction with the nanopowder surface. A magnetic stirrer can be used for stirring, with the stirring speed set at 500-800 rpm.

[0061] Dosage of silane coupling agent: Add 50g of silane coupling agent to the well dispersed nanopowder solution.

[0062] Combined with acrylic resin: (1) The grafted nanocomposite powder was separated by centrifugation and dried to obtain the processed nanoparticles.

[0063] (2) The grafted nanoparticles were mixed with acrylic resin (such as photosensitive resin) in acetone solvent. Acetone was chosen as the reaction solvent because it does not react with either material, but it can promote the bonding of chemical bonds and thus promote the reaction between the two materials. The materials were subjected to vacuum rotary evaporation at 35°C, with continuous rotation to ensure uniform mixing of the powder and resin. The stirring speed was set at 200-300 rpm, and the stirring time was set at 30-60 minutes. Finally, a photosensitive composite slurry with a certain degree of fluidity was obtained.

[0064] Powder content: The obtained grafted barium titanate-zirconium tungstate was mixed with 5 ml of acetone and 6 ml of UV-curable resin, and then vacuum-evaporated. The powder mass fraction was approximately 40 wt% to ensure sufficient flowability and printing performance.

[0065] DLP printing for the fabrication of flexible piezoelectric materials: (1) Pour the prepared photosensitive composite slurry into the liquid tank of the DLP (Digital Light Processing) 3D printer.

[0066] (2) Design the required 3D model on the computer and use slicing software to convert it into a file format suitable for DLP printers. Set the printing parameters of the DLP printer, such as layer thickness and curing time.

[0067] (3) Start the DLP printer and cure the composite paste layer by layer until the printing of the flexible piezoelectric material is complete. During the printing process, it is necessary to pay close attention to the printing quality to ensure that the printed part is consistent with the design model.

[0068] (4) Remove the printed flexible piezoelectric material from the liquid tank and clean it with anhydrous ethanol to remove any residual photosensitive resin. Gently brush the surface of the printed material with a soft brush to remove any uncured resin. Then immerse the printed material in anhydrous ethanol for about 15-30 minutes to thoroughly remove any residue.

[0069] (5) Remove the printed parts from the cleaning solution and dry them with compressed air. Then place the printed parts in a UV curing chamber for post-curing. Set the curing time to 30-60 minutes according to the actual situation. Fully cured printed parts have better mechanical properties and dimensional stability.

[0070] (6) The printed flexible piezoelectric material can be subjected to post-heat treatment, such as annealing or sintering, to adjust its piezoelectric properties. The specific heat treatment temperature and time need to be adjusted according to the thermal stability and expected performance of the material. (For example, annealing can be performed in the temperature range of 500-800℃, and the treatment time can be set at 1-2 hours. The flexible piezoelectric material after heat treatment will have better piezoelectric properties and stability.)

[0071] Layer Thickness: Set the printing layer thickness to 25 micrometers. Thinner layer thicknesses improve printing accuracy but increase printing time; thicker layer thicknesses shorten printing time but may reduce printing accuracy. UV Curing Time: Set the UV curing time, typically 12 seconds per layer, but adjust according to actual conditions. Longer UV curing times improve the material's mechanical properties but increase printing time; shorter UV curing times shorten printing time but may reduce the material's mechanical properties.

[0072] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for preparing a flexible piezoelectric material resistant to thermal expansion, characterized in that: Includes the following steps: S1: Prepare barium titanate nanoparticles and zirconium tungstate nanoparticles. Mix barium titanate nanoparticles and zirconium tungstate nanoparticles in different proportions to obtain mixed nanoparticles. Modify the mixed nanoparticles and then combine them with acrylic resin to obtain photosensitive composite slurries of barium titanate and zirconium tungstate in different proportions. S2: The optimal ratio of barium titanate and zirconium tungstate mixed nanopowders with different proportions obtained in step S1 is calculated using a random forest model and a differential evolution algorithm. S3: Based on the optimal ratio of barium titanate and zirconium tungstate obtained in step S2, the mixture is modified and then combined with acrylic resin to obtain a photosensitive composite slurry with the optimal ratio of barium titanate and zirconium tungstate. The algorithm calculation in step S2 includes the following steps: D1: Measure the coefficient of thermal expansion of the mixed nanopowders of barium titanate and zirconium tungstate with different proportions obtained in step S1, collect the coefficients of thermal expansion, and preprocess the collected data. D2: Divide the data obtained in step D1 into a training set and a test set; D3: Train the random forest model using the training set from step D2; D4: Evaluate the performance of the random forest model obtained in step D3 on the test set obtained in step D2; D5: On the trained random forest model, the optimal ratio of barium titanate and zirconium tungstate is obtained by using the differential evolution algorithm, with the minimum piezoelectric performance as a constraint.

2. The method for preparing a thermally resistant flexible piezoelectric material according to claim 1, characterized in that: It also includes printing the photosensitive composite paste obtained from S3 into sensors using DLP photopolymerization technology.

3. The method for preparing a thermally resistant flexible piezoelectric material according to claim 1, characterized in that: The preparation of barium titanate in step S1 includes the following steps: A1: Weigh out barium carbonate and titanium dioxide and put them into a ball mill jar. Add an appropriate amount of grinding media and deionized water, and ball mill the mixture to obtain mixed powder A. A2: Centrifuge, dry and sieve the mixed powder A to obtain a uniform precursor mixed powder A; A3: Precursor mixed powder A is placed in a high-temperature furnace for high-temperature solid-phase reaction to obtain barium titanate nanopowder.

4. The method for preparing a thermally resistant flexible piezoelectric material according to claim 3, characterized in that: The preparation of zirconium tungstate in step S1 includes the following steps: B1: Weigh tungsten trioxide and zirconium dioxide and put them into a ball mill jar. Add an appropriate amount of grinding media and deionized water, and ball mill the mixture to obtain mixed powder B. B2: The mixed powder B is centrifuged, dried and sieved to obtain a uniform precursor mixed powder B; B3: Precursor mixed powder B is placed in a high-temperature furnace for high-temperature solid-phase reaction to obtain zirconium tungstate nanopowder.

5. The method for preparing a flexible piezoelectric material resistant to thermal expansion according to claim 1, characterized in that: The modification processes in steps S1 and S3 include the following steps: C1: Place the prepared barium titanate nanopowder and zirconium tungstate nanopowder into a clean container and add anhydrous ethanol; C2: Add 3-(trimethoxysilyl)propyl methacrylate to the well-dispersed nanopowder solution; C3: After the grafting reaction is completed, the grafted nanoparticles are collected by centrifugation and drying; The combination of the mixed nanopowder and acrylic resin in steps S1 and S3 includes mixing the modified mixed nanopowder with acrylic resin in acetone and obtaining a photosensitive composite slurry with a certain fluidity by vacuum rotary evaporation.

6. The method for preparing a thermally resistant flexible piezoelectric material according to claim 1, characterized in that: The preprocessing in step D1 includes removing outliers from the collected data and normalizing the collected data; The step D2 of dividing the training set and the test set includes dividing the collected experimental data into a ratio of 80% training set and 20% test set. The training set is used for model training and the test set is used for model performance evaluation. Step D3 includes training the random forest model with the training set, calculating the loss value, setting the accuracy requirement of the model, and determining whether the accuracy requirement has been met. If the accuracy requirement is not met, the hyperparameters are adjusted and the random forest model is retrained. If the requirement is met, the trained random forest model is output. In step D4, the performance index of the model is determined by calculating the error between the predicted coefficient of thermal expansion and the actual value.

7. The method for preparing a flexible piezoelectric material resistant to thermal expansion according to claim 1, characterized in that: The differential evolution algorithm optimization in step D5 includes the following steps: E1: Initialize the population: Randomly generate a certain number of initial solutions as the initial population; E2: Selection operation: Use roulette wheel or other selection operators to select the next generation of the population; E3: Crossover operation: Generate new solutions according to a certain crossover strategy and add them to the next generation of the population; E4: Mutation operation: Calculates a new solution using the difference vector and updates the current solution to align with the new solution; E5: Determine the termination condition: Determine if the termination condition is met, such as the number of generations reaching the upper limit or the objective function value reaching the preset accuracy requirement; if the requirements are not met, continue to update the population, use the new population to replace the old population, and continue to execute step E2; if the requirements are met, output the result to obtain the optimal ratio of barium titanate and zirconium tungstate.

8. The method for preparing a flexible piezoelectric material resistant to thermal expansion according to claim 1, characterized in that: The process of printing composite materials into sensors using DLP photopolymerization technology includes the following steps: F1: Pour the prepared photosensitive composite paste into the liquid tank of the DLP 3D printer, and set the printing parameters of the DLP printer according to the designed 3D model; F2: Start the DLP printer and cure the photosensitive composite paste layer by layer until the printing of the flexible piezoelectric material is completed; F3: Remove the printed flexible piezoelectric material from the liquid tank, perform post-processing, and place the cleaned print into the UV curing chamber; F4: As needed, perform post-printed flexible piezoelectric material heat treatment to adjust its piezoelectric properties.

9. The flexible piezoelectric material obtained by the preparation method of a thermally resistant flexible piezoelectric material according to any one of claims 1-8, characterized in that: The material includes barium titanate, zirconium tungstate, and silane coupling agent; The silane coupling agent is 3-(trimethoxysilyl)propyl methacrylate.

Images