Intelligent decision-making method based on multi-mode bullet velocity measurement data

Abstract

The invention discloses an intelligent decision-making method based on multi-modal projectile velocity measurement data, and relates to the technical field of projectile velocity measurement, and the method comprises the steps of module deployment initialization, environment collection, synchronous triggering, adaptation degree calculation, weight fusion, error correction and decision-making storage. The method has the advantages that by deploying the multi-modal speed measurement module and the environment perception sub-module integrated with the multi-parameter detection function, the pre-trained environment-modal adaptation degree model is called to calculate the adaptation degree of each module and generate the dynamic priority list, and the data fusion weight is dynamically allocated according to the priority; the weight of an abnormal module is adjusted in combination with environmental interference conditions, and errors are compensated based on historical data of similar scenes; the defects that in the prior art, a single speed measurement module is difficult to adapt to complex environments and bullets with different attributes, data fusion weight fixing lacks scene pertinence, and error correction has no accurate basis, so that the speed measurement precision is low and the decision is unreliable are effectively overcome.

Description

Technical Field

[0001] This invention relates to the field of projectile velocity measurement technology, specifically to an intelligent decision-making method based on multimodal projectile velocity measurement data. Background Technology

[0002] Projectile velocity measurement technology is a core supporting technology for the performance verification of military weapon systems, the safety calibration of civilian ejection equipment (such as aviation life-saving ejection devices and industrial ejection testing), and ballistics research. Its velocity measurement accuracy and decision reliability directly determine the accuracy of equipment performance evaluation, the safety of use, and the effectiveness of subsequent parameter optimization. Existing intelligent decision-making methods based on multimodal projectile velocity measurement data mostly employ a single velocity measurement module, which is difficult to adapt to complex environments and projectiles with different properties. Furthermore, the data fusion weights are fixed and the error correction lacks precise basis, resulting in low velocity measurement accuracy and unreliable decisions. In addition, in existing technologies, velocity measurement data and environmental data are not collected synchronously, and the decision results lack traceability support. Historical data is scattered and difficult to reuse, leading to inefficient problem investigation and a lack of basis for solution iteration. Therefore, we propose an intelligent decision-making method based on multimodal projectile velocity measurement data. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent decision-making method based on multimodal projectile velocity measurement data.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an intelligent decision-making method based on multimodal projectile velocity measurement data, comprising the following steps: Step 1: Deploy and initialize the laser speed measurement module, electromagnetic induction speed measurement module, high-speed camera speed measurement module, and synchronization clock module, with a synchronization clock error of <1μs; Step 2: Deploy the environmental awareness submodule, set the environmental parameter collection threshold, and start real-time data collection; Step 3: Enter projectile attributes, capture launch trigger signal, and simultaneously trigger velocity measurement and environmental parameter acquisition; Step 4: Call the environment-modal adaptation model, input the adaptation degree between the environment and the projectile data calculation module, and generate a dynamic priority list; Step 5: Assign weights according to priority and then weight and fuse the speed measurement data; Step 6: Calculate data deviation, trigger conflict correction and compensate for error, and verify that the final speed error is <±1.5%; Step 7: Input the final speed into the intelligent decision-making model output and store all data.

[0005] As a further aspect of the present invention: In step one, the measurement accuracy of the laser velocity measurement module is preset to ±0.1%, and the velocity measurement range is set to Mach 1.2 to Mach 5, which is used to adapt to the velocity measurement of high-speed projectiles in clear, interference-free scenarios. The working environment parameters of the electromagnetic induction velocity measurement module are preset to a temperature range of -40℃ to 85℃ and a humidity range of 0-90%, and the velocity measurement range is set to Mach 0.3 to Mach 3, which is used to adapt to the velocity measurement of magnetic projectiles and in harsh temperature and humidity scenarios. The frame rate of the high-speed camera velocity measurement module is preset to 20,000 fps, the image resolution is set to 1920×1080, and the velocity measurement range is set to Mach 0.5 to Mach 2, which is used to adapt to the velocity measurement of non-metallic projectiles and in near-sonic scenarios. After deployment, all modules are initialized, including module power supply self-test and signal transmission link test, to ensure that there are no hardware faults in each module and that signal transmission is normal. After initialization, the module enters a standby state, waiting for the projectile launch signal to trigger.

[0006] As a further aspect of the present invention: In step two, the deployed environmental sensing submodule integrates a temperature and humidity sensor, an electromagnetic interference detector, and a Mach number calculator, and also integrates a beam attenuation detector to collect the beam attenuation rate of the laser velocity measurement module in real time. The collection frequency is consistent with the environmental parameters. The Mach number calculator calculates the estimated Mach number of the projectile by receiving preset parameters before the projectile is launched and the real-time air speed of sound. Then, it sets the environmental parameter collection thresholds: humidity > 70% is set as the "high humidity environment" threshold, humidity < 30% is set as the "low humidity environment" threshold, electromagnetic interference intensity > 50dB is set as the "strong electromagnetic interference environment" threshold, electromagnetic interference intensity < 20dB is set as the "weak electromagnetic interference environment" threshold, Mach number < 0.8 is set as the "low speed range" threshold, 0.8-1.2 Mach is set as the "near-sonic range" threshold, and > 1.2 Mach is set as the "high speed range" threshold. After the threshold settings are completed, the environmental sensing submodule is started to enter the real-time data collection state, and the collection frequency is set to 100Hz.

[0007] As a further aspect of the present invention: In step three, the projectile's attribute information includes the projectile's material, diameter, and mass. The projectile's attribute information is stored in a local database for subsequent modal adaptation calculation. When the projectile launching device issues a launch signal, the launch trigger signal is captured by a signal sensor and synchronously transmitted to the multimodal velocity measurement module and environmental perception submodule in step one. This triggers the multimodal velocity measurement module to start collecting velocity data during the projectile's flight, while simultaneously triggering the environmental perception submodule to record the trigger time and real-time environmental parameters within the following 500ms. Both the collected velocity measurement data and environmental parameters carry the timestamp of the synchronization clock module.

[0008] As a further aspect of the present invention: In step four, a pre-trained "environment-modal fit model" is called from the local database. This model is generated through training on 100,000 sets of historical "environmental parameters-modal errors" data. The training process uses a gradient descent algorithm to optimize the model parameters, and the model output is the fit value of each speed measurement module. First, the real-time environmental parameters captured in step three are extracted and denoted as... Including humidity Electromagnetic interference intensity and Mach number Bullet attribute information, denoted as Including materials ,diameter ,Will and Input the data into the environment-modal fitness model and calculate the fitness value for each velocity measurement module: Laser velocimetry module compatibility value : ; in, The weighting for humidity is set to 0.4. The weight for the impact of electromagnetic interference is set to 0.3. The weight for Mach number influence is set to 0.3, and the projectile material is... It is applicable to both magnetic metals and non-metals. >70%, When any condition >50% is met, , The weights were adjusted to 0.6 and 0.4 respectively; Electromagnetic induction speed measurement module compatibility value : ; in, The weighting for the effect of temperature is set to 0.5. The weight for electromagnetic interference is set to 0.5, only applicable to the projectile material. It is a magnetic metal and effective at temperatures ranging from -40℃ to 85℃. When it is non-metallic Automatically set to 0, Real-time ambient temperature; High-speed camera speed measurement module compatibility value : ; in, The weighting for humidity is set to 0.3. The weight for Mach number influence is set to 0.7, and the projectile material is... It is applicable to both magnetic metals and non-metals. When in the Mach 0.8-1.2 range, The weight was adjusted to 0.8; Based on calculations , and The numerical values ​​are used to prioritize the three types of speed measurement modules. The ranking rule is that the higher the fit value, the higher the priority. If the fit value of a module is 0, it is excluded from the priority ranking. Finally, a dynamic priority list of multimodal speed measurement modules for this test scenario is generated.

[0009] As a further aspect of the present invention: In step five, according to the dynamic priority list, data fusion weights are assigned to the speed measurement modules corresponding to each priority: the speed measurement module weight for priority 1... The weight is set to 0.6-0.8 (the specific value is determined based on the difference between the adaptation value and the adaptation values ​​of other modules; 0.8 is used when the difference is >0.2, 0.7 is used when the difference is 0.1-0.2, and 0.6 is used when the difference is <0.1), which is the weight of the speed measurement module with priority 2. Set to 0.1-0.3 (values ​​range from 1- - ,in (Priority 3 weight), the weight of the speed measurement module with priority 3. Set to 0-0.2 (enabled only when three types of valid modules exist; otherwise, if only two types of valid modules exist...) ), and must meet Subsequently, the velocity measurement data from each module collected in step three, and the laser module data, were extracted. Electromagnetic module data and high-speed camera module data For valid data (module data with a fit value ≠ 0), a weighted fusion calculation is performed according to the following formula: ; After fusion is complete, store the fusion speed value. And the corresponding weight allocation records.

[0010] As a further aspect of the present invention: in step six, the fusion speed value is extracted. The system retrieves the original data from each module and calculates the deviation between any two valid module data. When the deviation is greater than ±2%, conflict correction is triggered. First, it combines environmental parameters to determine interference (when the laser module beam attenuation rate is greater than 30% or the electromagnetic module is subjected to strong interference greater than 50dB, the weight of the corresponding module is reduced to 0.3 and 0.2, and the weight of other modules is increased). Then, it calls 50,000 sets of historical valid data from similar scenarios and calculates the error compensation value (the difference between the average historical fusion value and the average historical original data). This value is then added to the current fusion value to obtain the final speed value. The error of the final speed value is verified and must be less than ±1.5%. If it fails, the correction is repeated until the target is met.

[0011] As a further aspect of the present invention: In step seven, the final speed value is input into the intelligent decision-making model, and the result is output according to the application scenario. Projectile performance testing: If the final velocity value is within the preset standard range and the difference between the final velocities of three consecutive tests is less than ±3m / s, it is considered qualified; otherwise, it is unqualified. Launch parameter optimization: If the final velocity is more than 10% lower than the target velocity, it is recommended to increase the propellant charge; if the final velocity is more than 5% higher than the target velocity, it is recommended to reduce the propellant charge. Fault diagnosis: The final speed value drops sharply by more than 10% compared to the final speed of the previous test, and there is no environmental interference. ≤70%, ≤50%, beam attenuation rate ≤30%), indicating "barrel wear / projectile center of gravity shift"; The decision results are displayed through a human-computer interface, and all test data are stored in JSON format in a local database and in the cloud to ensure traceability.

[0012] Compared with the prior art, the beneficial effects of the present invention by adopting the above technical solution are as follows: 1. This invention deploys a multimodal velocity measurement module and an environmental perception submodule that integrates multi-parameter detection functions. It simultaneously captures projectile attributes, real-time environmental data, and velocity measurement data. It calls a pre-trained environment-modal adaptability model to calculate the adaptability of each module and generate a dynamic priority list. It dynamically allocates data fusion weights according to priority, and adjusts the weights of abnormal modules based on environmental interference. It also compensates for errors based on historical data from similar scenarios. This effectively solves the defects in existing technologies, such as the difficulty of adapting a single velocity measurement module to complex environments and projectiles with different attributes, the lack of scenario-specificity in fixed data fusion weights, and the lack of precise basis for error correction, which lead to low velocity measurement accuracy and unreliable decision-making. Ultimately, it achieves accurate measurement of projectile velocity in complex scenarios and improves the reliability of velocity measurement data. 2. This invention synchronously triggers the multimodal velocity measurement module and the environmental perception submodule to collect data and attach a unified timestamp. The final velocity value is then linked with the scenario data and input into the intelligent decision-making model. This enables scenario-based output for projectile performance testing, launch parameter optimization, and fault diagnosis. Simultaneously, all test data is synchronously stored locally and in the cloud in a standardized format. This effectively solves the shortcomings of existing technologies, such as asynchronous collection of velocity measurement data and environmental data, lack of data traceability support for decision results, fragmented and unusable historical data leading to inefficient problem investigation, and lack of basis for scheme iteration. Ultimately, it improves the data traceability and problem location efficiency of the projectile testing process, providing complete data support for subsequent velocity measurement model optimization and test scheme adjustment in different scenarios. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the method flow in an embodiment of the present invention. Detailed Implementation

[0014] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0015] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0016] Please see the appendix Figure 1 This invention discloses an intelligent decision-making method based on multimodal projectile velocity measurement data, comprising the following steps: Step 1: Deploy and initialize the laser speed measurement module, electromagnetic induction speed measurement module, high-speed camera speed measurement module, and synchronization clock module, with a synchronization clock error of <1μs; Step 2: Deploy the environmental awareness submodule, set the environmental parameter collection threshold, and start real-time data collection; Step 3: Enter projectile attributes, capture launch trigger signal, and simultaneously trigger velocity measurement and environmental parameter acquisition; Step 4: Call the environment-modal adaptation model, input the adaptation degree between the environment and the projectile data calculation module, and generate a dynamic priority list; Step 5: Assign weights according to priority and then weight and fuse the speed measurement data; Step 6: Calculate data deviation, trigger conflict correction and compensate for error, and verify that the final speed error is <±1.5%; Step 7: Input the final speed into the intelligent decision-making model output and store all data.

[0017] In one embodiment of the present invention: In step one, the measurement accuracy of the laser velocity measurement module is preset to ±0.1%, and the velocity measurement range is set to Mach 1.2 to Mach 5, which is used to adapt to the velocity measurement of high-speed projectiles in clear, interference-free scenarios. The working environment parameters of the electromagnetic induction velocity measurement module are preset to a temperature range of -40℃ to 85℃ and a humidity range of 0-90%, and the velocity measurement range is set to Mach 0.3 to Mach 3, which is used to adapt to the velocity measurement of magnetic projectiles and in harsh temperature and humidity scenarios. The frame rate of the high-speed camera velocity measurement module is preset to 20,000fps, the image resolution is set to 1920×1080, and the velocity measurement range is set to Mach 0.5 to Mach 2, which is used to adapt to the velocity measurement of non-metallic projectiles and in near-sonic scenarios. After deployment, all modules are initialized, including module power supply self-test and signal transmission link test, to ensure that there are no hardware faults in each module and that signal transmission is normal. After initialization, the module enters a standby state, waiting for the projectile launch signal to trigger.

[0018] In one embodiment of the present invention: In step two, the deployed environmental sensing submodule integrates a temperature and humidity sensor, an electromagnetic interference detector, and a Mach number calculator, and also integrates a beam attenuation detector to collect the beam attenuation rate of the laser velocity measurement module in real time. The collection frequency is consistent with the environmental parameters. The Mach number calculator calculates the estimated Mach number of the projectile by receiving preset parameters before the projectile is launched and the real-time air speed of sound. Then, it sets the environmental parameter collection thresholds: humidity > 70% is set as the "high humidity environment" threshold, humidity < 30% is set as the "low humidity environment" threshold, electromagnetic interference intensity > 50dB is set as the "strong electromagnetic interference environment" threshold, electromagnetic interference intensity < 20dB is set as the "weak electromagnetic interference environment" threshold, Mach number < 0.8 is set as the "low speed range" threshold, 0.8-1.2 Mach is set as the "near-sonic range" threshold, and > 1.2 Mach is set as the "high speed range" threshold. After the threshold settings are completed, the environmental sensing submodule is started to enter the real-time data collection state, and the collection frequency is set to 100Hz.

[0019] In one embodiment of the present invention: In step three, the projectile's attribute information includes the projectile material, projectile diameter, and projectile mass. The projectile attribute information is stored in a local database for subsequent modal adaptation calculation. When the projectile launching device sends a launch signal, the launch trigger signal is captured by a signal sensor and synchronously transmitted to the multimodal velocity measurement module and environmental perception submodule in step one. This triggers the multimodal velocity measurement module to start collecting velocity data during the projectile's flight process, and simultaneously triggers the environmental perception submodule to record the trigger time and real-time environmental parameters within the following 500ms. The collected velocity measurement data and environmental parameters both carry the timestamp of the synchronization clock module.

[0020] In one embodiment of the present invention: In step four, a pre-trained "environment-modal fit model" is called from the local database. This model is generated through training on 100,000 sets of historical "environmental parameters-modal errors" data. The training process uses a gradient descent algorithm to optimize the model parameters. The model output is the fit value of each speed measurement module. First, the real-time environmental parameters captured in step three are extracted and denoted as... Including humidity Electromagnetic interference intensity and Mach number Bullet attribute information, denoted as Including materials ,diameter ,Will and Input the data into the environment-modal fitness model and calculate the fitness value for each velocity measurement module: Laser velocimetry module compatibility value : ; in, The weighting for humidity is set to 0.4. The weight for the impact of electromagnetic interference is set to 0.3. The weight for Mach number influence is set to 0.3, and the projectile material is... It is applicable to both magnetic metals and non-metals. >70%, When any condition >50% is met, , The weights were adjusted to 0.6 and 0.4 respectively; Electromagnetic induction speed measurement module compatibility value : ; in, The weighting for the effect of temperature is set to 0.5. The weight for electromagnetic interference is set to 0.5, only applicable to the projectile material. It is a magnetic metal and effective at temperatures ranging from -40℃ to 85℃. When it is non-metallic Automatically set to 0, Real-time ambient temperature; High-speed camera speed measurement module compatibility value : ; in, The weighting for humidity is set to 0.3. The weight for Mach number influence is set to 0.7, and the projectile material is... It is applicable to both magnetic metals and non-metals. When in the Mach 0.8-1.2 range, The weight was adjusted to 0.8; Based on calculations , and The numerical values ​​are used to prioritize the three types of speed measurement modules. The ranking rule is that the higher the fit value, the higher the priority. If the fit value of a module is 0, it is excluded from the priority ranking. Finally, a dynamic priority list of multimodal speed measurement modules for this test scenario is generated.

[0021] In one embodiment of the present invention: In step five, according to the dynamic priority list, data fusion weights are assigned to the speed measurement modules corresponding to each priority: the speed measurement module weight for priority 1... The weight is set to 0.6-0.8 (the specific value is determined based on the difference between the adaptation value and the adaptation values ​​of other modules; 0.8 is used when the difference is >0.2, 0.7 is used when the difference is 0.1-0.2, and 0.6 is used when the difference is <0.1), which is the weight of the speed measurement module with priority 2. Set to 0.1-0.3 (values ​​range from 1- - ,in (Priority 3 weight), the weight of the speed measurement module with priority 3. Set to 0-0.2 (enabled only when three types of valid modules exist; otherwise, if only two types of valid modules exist...) ), and must meet Subsequently, the velocity measurement data from each module collected in step three, and the laser module data, were extracted. Electromagnetic module data and high-speed camera module data For valid data (module data with a fit value ≠ 0), a weighted fusion calculation is performed according to the following formula: ; After fusion is complete, store the fusion speed value. And the corresponding weight allocation records.

[0022] In one embodiment of the present invention: in step six, the fusion speed value is extracted. The system retrieves the original data from each module and calculates the deviation between any two valid module data. When the deviation is greater than ±2%, conflict correction is triggered. First, the system combines environmental parameters to determine the interference, then calls 50,000 sets of historical valid data from similar scenarios to calculate the error compensation value. This value is then superimposed on the current fusion value to obtain the final speed value. The error of the final speed value is verified and must be less than ±1.5%. If it fails, the correction is repeated until the target is met.

[0023] In one embodiment of the present invention: In step seven, the final speed value is input into the intelligent decision-making model, and the result is output according to the application scenario: Projectile performance testing: If the final velocity value is within the preset standard range and the difference between the final velocities of three consecutive tests is less than ±3m / s, it is considered qualified; otherwise, it is unqualified. Launch parameter optimization: If the final velocity is more than 10% lower than the target velocity, it is recommended to increase the propellant charge; if the final velocity is more than 5% higher than the target velocity, it is recommended to reduce the propellant charge. Fault diagnosis: The final velocity value drops by more than 10% compared to the final velocity of the previous test, and there is no environmental interference, indicating "barrel wear / projectile center of gravity shift"; The decision results are displayed through a human-computer interface, and all test data are stored in JSON format in a local database and in the cloud to ensure traceability.

[0024] Example 1, please refer to the appendix. Figure 1 Decision-making for high-speed magnetic projectile velocity measurement under clear, low-humidity, and weak-interference conditions: 1. Test Scenario and Parameter Settings Projectile properties: Material is magnetic alloy steel, diameter is 12mm, weight is 15g, stored in local database; Environmental parameters: Humidity 25% (low humidity, <30%), electromagnetic interference 15dB (weak interference, <20dB), temperature 28℃, estimated projectile Mach number 2.0 (high speed, >1.2 Mach). Module initialization: Laser velocimetry module (accuracy ±0.1%, range 1.2-5 Mach), electromagnetic induction module (-40-85℃, 0.3-3 Mach), high-speed camera (20000fps, 0.5-2 Mach), synchronization clock error 0.8. .

[0025] 2. Fit Calculation Laser module compatibility According to the formula , =0.4、 =0.3、 =0.3, substitute =25、 =15、 =2.0, therefore ; Electromagnetic module compatibility According to the formula , =0.5、 =0.5, substitute =28、 =15, therefore... =0.5×0.95+0.5×0.85=0.9; High-speed camera compatibility According to the formula , =0.3、 =0.7, substitute =25、 =2.0, therefore ≈0.341.

[0026] 3. Weighting and Data Fusion Priority: (0.9) > (0.675) > (0.341), =0.8 (difference 0.225 > 0.2) =0.2、 =0; Speed ​​measurement data: = Mach 2.0 = Mach 1.98 = Mach 1.95, fusion =0.8×1.98+0.2×2.0=1.984 Mach.

[0027] 4. Error Correction and Decision Making Data deviation of 1.01% < 2%, no correction required; final speed error of 1.2% < 1.5%. Decision: The projectile performance is satisfactory (the difference in three consecutive tests is less than 3 m / s), and the launch parameters do not need to be adjusted.

[0028] Example 2, please refer to the appendix. Figure 1 Velocity measurement decision for non-metallic near-sonic projectiles under high humidity and strong interference: 1. Test Scenario and Parameter Settings Projectile properties: Material: carbon fiber (non-metallic), diameter: 10mm, weight: 8g; Environmental parameters: Humidity 80% (high humidity, >70%), electromagnetic interference 60dB (strong interference, >50dB), temperature 30℃, Mach number 1.0 (near-sonic). Module status: Electromagnetic module =0 (non-metallic), only effective for lasers and high-speed cameras.

[0029] 2. Fit Calculation laser module : =0.6、 =0.4 (H>70%), substitute into =80、 =60、 =1.0, =0.6×0.2+0.4×0.4+0.3×0.2=0.34; high-speed camera : =0.8 (nearsonic speed), substitute into =80、 =1.0, =0.3×0.2+0.8×1=0.86.

[0030] 3. Weighting and Data Fusion Priority: (0.86) > (0.34), =0.8、 =0.2; Speed ​​measurement data: = Mach 1.02 =Mach 1.0, fusion =0.8×1.0+0.2×1.02=1.004 Mach.

[0031] 4. Error Correction and Decision Making A deviation of 2% requires no correction; an error of 1.0% is less than 1.5%. Decision: The projectile performance is satisfactory, with no indication of barrel wear or center of gravity shift.

[0032] By deploying a multimodal velocity measurement module and an environmental perception submodule that integrates temperature and humidity, electromagnetic interference, and beam attenuation detection, projectile attributes and real-time environmental / velocity data are collected synchronously. A pre-trained environment-modal adaptability model is invoked and a dynamic priority list is generated. Data fusion weights are allocated according to adaptability differences. The weights of abnormal modules are adjusted in conjunction with environmental interference and historical data error compensation for similar scenarios. Finally, the speed and scenario data are linked and input into an intelligent decision-making model. All data is stored in a standardized manner locally and in the cloud. This effectively solves the defects of existing technologies, such as poor adaptability of a single velocity measurement module, fixed weights, lack of basis for error correction, data asynchrony, difficulty in tracing, and fragmented historical data.

[0033] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Any variations and modifications can be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention, without departing from the scope of the invention, fall within the protection scope defined by the claims of the present invention.

Claims

1. An intelligent decision-making method based on multimodal projectile velocity measurement data, comprising an intelligent decision-making method, characterized in that, The intelligent decision-making method includes the following steps: Step 1: Deploy and initialize the laser speed measurement module, electromagnetic induction speed measurement module, high-speed camera speed measurement module, and synchronization clock module, with a synchronization clock error of <1μs; Step 2: Deploy the environmental awareness submodule, set the environmental parameter collection threshold, and start real-time data collection; Step 3: Enter projectile attributes, capture launch trigger signal, and simultaneously trigger velocity measurement and environmental parameter acquisition; Step 4: Call the environment-modal adaptation model, input the adaptation degree between the environment and the projectile data calculation module, and generate a dynamic priority list; Step 5: Assign weights according to priority and then weight and fuse the speed measurement data; Step 6: Calculate data deviation, trigger conflict correction and compensate for error, and verify that the final speed error is <±1.5%; Step 7: Input the final speed into the intelligent decision-making model output and store the full data.

2. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step one, the measurement accuracy of the laser velocity measurement module is preset to ±0.1%, and the velocity measurement range is set to Mach 1.2 to Mach 5. The working environment parameters of the electromagnetic induction velocity measurement module are preset to a temperature range of -40℃ to 85℃ and a humidity range of 0-90%, with a velocity measurement range of Mach 0.3 to Mach 3. The frame rate of the high-speed camera velocity measurement module is preset to 20000fps, the image resolution is set to 1920×1080, and the velocity measurement range is set to Mach 0.5 to Mach 2.

3. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step two, the deployed environmental sensing submodule integrates a temperature and humidity sensor, an electromagnetic interference detector, and a Mach number calculator. It also integrates a beam attenuation detector to collect the beam attenuation rate of the laser velocity measurement module in real time. The collection frequency is consistent with the environmental parameters. The Mach number calculator calculates the estimated Mach number of the projectile by receiving preset parameters before the projectile is launched and the real-time air speed of sound. Then, it sets the environmental parameter collection thresholds: humidity > 70% is set as the "high humidity environment" threshold, humidity < 30% is set as the "low humidity environment" threshold, electromagnetic interference intensity > 50dB is set as the "strong electromagnetic interference environment" threshold, electromagnetic interference intensity < 20dB is set as the "weak electromagnetic interference environment" threshold, Mach number < 0.8 is set as the "low speed range" threshold, 0.8-1.2 Mach is set as the "near-sonic range" threshold, and > 1.2 Mach is set as the "high speed range" threshold. After the threshold settings are completed, the environmental sensing submodule is started to enter the real-time data collection state, and the collection frequency is set to 100Hz.

4. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step three, the projectile's attribute information includes the projectile's material, diameter, and mass. This projectile attribute information is stored in a local database for subsequent modal adaptation calculations. When the projectile launching device emits a launch signal, the signal sensor captures the launch trigger signal and synchronously transmits it to the multimodal velocity measurement module and environmental perception submodule in step one. This triggers the multimodal velocity measurement module to start collecting velocity data during the projectile's flight, while simultaneously triggering the environmental perception submodule to record the trigger time and real-time environmental parameters within the following 500ms. Both the collected velocity measurement data and environmental parameters carry the timestamp from the synchronization clock module.

5. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step four, the pre-trained "environment-modal fit model" in the local database is invoked. This model is generated through training on 100,000 sets of historical "environmental parameter-modal error" data. The training process uses the gradient descent algorithm to optimize the model parameters. The model output is the fit value of each speed measurement module. First, the real-time environmental parameters captured in step three are extracted and denoted as... Including humidity Electromagnetic interference intensity and Mach number Bullet attribute information, denoted as Including materials ,diameter ,Will and Input the data into the environment-modal fitness model and calculate the fitness value for each velocity measurement module: Laser velocimetry module compatibility value : ； in, The weighting for humidity is set to 0.

4. The weight for the impact of electromagnetic interference is set to 0.

3. The weight for Mach number influence is set to 0.3, and the projectile material is... It is applicable to both magnetic metals and non-metals. >70%, When any condition >50% is met, , The weights were adjusted to 0.6 and 0.4 respectively; Electromagnetic induction speed measurement module compatibility value : ； in, The weighting for the effect of temperature is set to 0.

5. The weight for electromagnetic interference is set to 0.5, only applicable to the projectile material. It is a magnetic metal and effective at temperatures ranging from -40℃ to 85℃. When it is non-metallic Automatically set to 0, Real-time ambient temperature; High-speed camera speed measurement module compatibility value : ； in, The weighting for humidity is set to 0.

3. The weight for Mach number influence is set to 0.7, and the projectile material is... It is applicable to both magnetic metals and non-metals. When in the Mach 0.8-1.2 range, The weight was adjusted to 0.8; Based on calculations , and The numerical values ​​are used to prioritize the three types of speed measurement modules. The ranking rule is that the higher the fit value, the higher the priority. If the fit value of a module is 0, it is excluded from the priority ranking. Finally, a dynamic priority list of multimodal speed measurement modules for this test scenario is generated.

6. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step five, data fusion weights are assigned to the speed measurement modules corresponding to each priority level according to the dynamic priority list: the speed measurement module weight for priority 1 is... The weight of the speed measurement module is set to 0.6-0.8, with a priority of 2. The weight of the speed measurement module is set to 0.1-0.3, with a priority of 3. Set to 0-0.2, and it must meet the following requirements. Subsequently, the velocity measurement data from each module collected in step three, and the laser module data, were extracted. Electromagnetic module data and high-speed camera module data The valid data is weighted and fused according to the following formula: ； After fusion is complete, store the fusion speed value. And the corresponding weight allocation records.

7. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step six, the fusion speed value is extracted. The system retrieves the original data from each module and calculates the deviation between any two valid module data. When the deviation is greater than ±2%, conflict correction is triggered. First, the system combines environmental parameters to determine the interference, then calls 50,000 sets of historical valid data from similar scenarios to calculate the error compensation value. This value is then superimposed on the current fusion value to obtain the final speed value. The error of the final speed value is verified and must be less than ±1.5%. If it fails, the correction is repeated until the target is met.

8. The intelligent decision-making method based on multimodal projectile velocity measurement data according to claim 1, characterized in that: In step seven, the final speed value is input into the intelligent decision-making model, and the results are output according to the application scenario: Projectile performance testing: If the final velocity value is within the preset standard range and the difference between the final velocities of three consecutive tests is less than ±3m / s, it is considered qualified; otherwise, it is unqualified. Launch parameter optimization: If the final velocity is more than 10% lower than the target velocity, it is recommended to increase the propellant charge; if the final velocity is more than 5% higher than the target velocity, it is recommended to reduce the propellant charge. Fault diagnosis: The final velocity value drops by more than 10% compared to the final velocity of the previous test, and there is no environmental interference, indicating "barrel wear / projectile center of gravity shift"; The decision results are displayed through a human-computer interface, and all test data are stored in JSON format in a local database and in the cloud to ensure traceability.

Images