A rocket core stage flight status monitoring micromodule and its design method

Abstract

The present invention relates to the field of aviation technology, and specifically to a rocket core sub-stage flight status monitoring micromodule and a design method thereof. The micromodule includes a power supply, a sampling circuit, a hot-plug connector, a power input protection circuit, a plug detection circuit and a control circuit; the power supply supplies power to the hot-plug connector through the power input protection circuit, and the power input protection circuit is used to control the connection and disconnection of the power supply; the plug detection circuit is used to detect the plug status of the hot-plug connector and send it to the control circuit; the control circuit includes a microprocessor, and the microprocessor is provided with a TensorFlow Lite for Microcontrollers module and an NNabla Embedded module; the present invention is simple to install, has high versatility, low cost and small size.

Description

Technical Field

[0001] The present invention relates to the field of aviation technology, and in particular to a rocket core-stage flight status monitoring micromodule and a design method thereof. Background Art

[0002] During the launch of a rocket and its sub-stage operation, condition monitoring is one of the key technologies to ensure mission safety and success. Condition monitoring includes but is not limited to monitoring the working conditions of a single machine, various environmental conditions, and the supply and consumption of electrical energy. Accurate condition monitoring helps the control system adjust the flight strategy in real time and respond promptly to various abnormal situations that may occur.

[0003] In the prior art, the status monitoring modules for rockets are usually developed by third-party stand-alone outsourcing manufacturers and are designed for specific models of rockets or specific mission requirements. They tend to have high reliability and accuracy, but also face disadvantages such as high cost, heavy weight, large size, and difficulty in generalization. In addition, the maintenance and upgrading of these systems are also relatively difficult, because each change may require the entire system to be re-verified. How to design a more economical, easy to integrate and widely applicable flight status monitoring micromodule for different rocket models, so that it can quickly adapt to a variety of mission requirements and board configurations, this is a problem that needs to be solved. Summary of the invention

[0004] In order to solve the problems of high cost, heavy weight and large volume of the status monitoring module in the prior art, the present invention proposes a rocket core-stage flight status monitoring micromodule and a design method thereof, so as to significantly reduce the weight, volume and manufacturing cost of the status monitoring micromodule.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a rocket core-stage flight status monitoring micromodule, including a power supply, a sampling circuit, a hot-swap connector, a power input protection circuit, a plug-in detection circuit and a control circuit;

[0006] The power supply supplies power to the hot-swappable connector through the power input protection circuit, the power input protection circuit uses an N-channel MOSFET to control the connection and disconnection of the power supply, and the gate of the MOSFET is provided with an RC delay circuit to achieve a smooth rise in current and avoid surge current;

[0007] The plug-in detection circuit includes a resistor voltage divider network and a comparator, which is used to detect the plug-in status of the hot-plug connector and send it to the control circuit;

[0008] The control circuit includes a microprocessor, which is used to read the plug-in detection signal to determine the plug-in state and control the MOSFET according to the plug-in state; it is also used to read the voltage drop value on the current detection resistor to monitor the system working current during the micromodule plug-in process; the microprocessor is provided with a TensorFlow Lite for Microcontrollers module and an NNablaEmbedded module;

[0009] Among them, the TensorFlow Lite for Microcontrollers module is used to perform basic model reasoning according to the data input from the sensor, evaluate the basic state of the rocket core stage through basic model reasoning, and output the basic reasoning results of the state of the rocket core stage; the NNabla Embedded module is used to output the complex reasoning results of the state of the rocket core stage through the neural network model based on the basic reasoning results and the original sensor data when the basic reasoning results exceed the preset threshold.

[0010] The microprocessor is also provided with a fusion decision module, which is used to fuse the basic reasoning results with the complex reasoning results, make a final decision based on the fusion results, and trigger corresponding control operations according to the final decision.

[0011] The fusion decision module fuses the basic reasoning results with the complex reasoning results by weighted averaging.

[0012] The control operations triggered by the fusion decision module include: adjusting attitude control and / or sending a flight alert.

[0013] The microprocessor is also provided with a data acquisition and preprocessing module, and the data acquisition and preprocessing module is used to preprocess the collected data to meet the requirements of the neural network input in the NNabla Embedded module.

[0014] The software algorithm in the microprocessor is designed with a layered architecture, including:

[0015] Data acquisition layer: responsible for acquiring data from sensors and performing preliminary filtering and preprocessing;

[0016] Basic reasoning layer: Use TensorFlow Lite for Microcontrollers for simple and efficient model reasoning;

[0017] Complex computing layer: Use NNabla Embedded to perform complex computing and optimization tasks;

[0018] Decision-making layer: Combines the results of basic reasoning and complex calculations to perform status monitoring and decision-making.

[0019] In addition, the present invention also provides a design method for a rocket core-stage flight status monitoring micromodule, comprising the following steps:

[0020] Step 1: Determine the comprehensive cost target, volume and weight requirements of the module;

[0021] Step 2: Determine the power supply parameters, monitoring parameters and environmental adaptation parameters of the module;

[0022] Step 3: Select and design circuit components based on comprehensive cost targets, power supply parameters, monitoring parameters, and environmental adaptability parameters;

[0023] Step 4: Select and design a connector suitable for hot plugging, and cooperate with the power input protection circuit and plug detection circuit to ensure that the microcontroller can sense the change of connection status during the hot plugging process;

[0024] Step 5: Based on the selection of circuit components, design and draw the circuit schematic; at the same time, based on the volume and weight requirements of the module, perform PCB layout design;

[0025] Step 6: Architect the software algorithm for the microprocessor.

[0026] In step 1, the calculation formula of the comprehensive cost target is:

[0027] C acc =C e +C d / N+C u ;

[0028] Among them, C acc represents the comprehensive cost, C e represents the development cost, C d represents the development cost, C u represents the mass production cost, and N represents the mass production quantity.

[0029] In step 2, the parameters of the power supply include input voltage range and power supply voltage;

[0030] Monitoring parameters include the number of monitoring channels and the single-channel sampling rate; environmental adaptation parameters include temperature range and insulation resistance value.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. The present invention solves the problems of high cost, heavy weight and large volume of the state monitoring module in the prior art: on the basis of meeting the general requirements of rocket core-level state monitoring, the software architecture of the microcontroller in the micromodule is improved to reduce the hardware requirements of the micromodule for the microprocessor. The present invention can significantly reduce the manufacturing cost of the state monitoring micromodule. The size of the micromodule designed by the present invention is similar to the 2019 version of the fifth set of 1-cent coins, and the weight of a single one does not exceed 5g. The standardized micromodule can achieve large-scale batch production, further reducing the production cost of a single module, and better adapting to the low-cost industry demand of the rapid development of commercial aerospace.

[0033] 2. The present invention can reduce the difficulty of installing the status monitoring module: the present invention supports hot plugging and hot backup, and the micromodules can be quickly installed and replaced, simplifying the installation and maintenance process of the entire monitoring system. The plug-in interface design allows non-professionals to easily perform system maintenance and troubleshooting.

[0034] 3. The present invention can improve the versatility of the state monitoring module: The micromodule of the present invention adopts a universal interface and protocol to ensure that the micromodule can be used on rockets of different models and designs. This versatility not only reduces the customization cost of the micromodule, but also improves the adaptability and flexibility of the entire system. At the same time, it also realizes complex state monitoring, anomaly detection and predictive maintenance in flight scenarios, and improves the intelligence level of the rocket. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A schematic structural diagram of a rocket core-stage flight status monitoring micromodule provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0037] Embodiment 1

[0038] like Figure 1As shown, an embodiment of the present invention provides a rocket core sub-stage flight status monitoring micromodule, including a power supply, a sampling circuit, a hot-plug connector, a power input protection circuit, a plug detection circuit and a control circuit; the power supply supplies power to the hot-plug connector through the power input protection circuit, and the power input protection circuit uses an N-channel MOSFET to control the connection and disconnection of the power supply, and the gate of the MOSFET is provided with an RC delay circuit to achieve a smooth rise in current and avoid surge current; the plug detection circuit includes a resistor divider network and a comparator, which are used to detect the plug status of the hot-plug connector and send it to the control circuit; the control circuit includes a microprocessor, which is used to read the plug detection signal to determine the plug status, and control the MOSFET according to the plug status; it is also used to read the voltage drop value on the current detection resistor to monitor the system working current during the micromodule plug-in process; the microprocessor is provided with a TensorFlow Litefor Microcontrollers module and a NNabla Embedded module.

[0039] Among them, the TensorFlow Lite for Microcontrollers module is used to perform basic model reasoning according to the data input from the sensor, evaluate the basic state of the rocket core stage through basic model reasoning, and output the basic reasoning results of the state of the rocket core stage; the NNabla Embedded module is used to output the complex reasoning results of the state of the rocket core stage through the neural network model based on the basic reasoning results and the original sensor data when the basic reasoning results exceed the preset threshold.

[0040] Specifically, TensorFlow Lite for Microcontrollers (TFLM) is a subset of TensorFlow Lite that supports a variety of common neural network layers and operations, is optimized for microcontrollers and other extremely low-power devices, and is suitable for devices with very limited resources. It has cross-platform support functions, including existing mainstream architectures such as ARM Cortex-M, RISC-V, and ESP32, which improves the versatility of the monitoring module. It is widely used in various embedded devices and low-power sensor nodes for machine learning reasoning. In this embodiment, it is mainly used to perform basic model reasoning for real-time monitoring of the status of rocket core stages. Due to its light weight and high efficiency, TFLM can run on resource-constrained embedded devices to achieve fast-response state monitoring. It quickly evaluates the basic state of the rocket core stage through sensor data input, and outputs preliminary evaluation results to achieve fast-response state monitoring.

[0041] NNabla Embedded: NNabla is Sony's open source deep learning framework, and its Embedded version is specially optimized for micro-embedded systems. It provides a wealth of APIs and tools for model training, optimization and deployment. It has highly optimized performance and is suitable for running on embedded devices. NNabla Embedded has a neural network model that can handle complex problems, thereby providing more accurate and in-depth analysis. In an embodiment of the present invention, NNabla Embedded is used to perform complex model reasoning, that is, to further analyze the situation when the basic reasoning results are abnormal. Specifically, when the basic model reasoning result of the TFLM module exceeds the preset threshold, NNabla Embedded is triggered to perform complex reasoning and optimization calculations. At this time, it will receive the basic reasoning results and the original sensor data, perform more complex reasoning and optimization calculations, and provide more accurate state evaluation. Therefore, in this embodiment, the basic reasoning results and complex reasoning results can be fused through the TensorFlow Lite for Microcontrollers module and the NNabla Embedded module to provide a more accurate state evaluation. Finally, the final decision is made based on the fusion result, and the corresponding control operation (such as adjusting attitude control, sending flight alarms, etc.) is triggered.

[0042] Furthermore, in this embodiment, a fusion decision module is also provided in the microprocessor, and the fusion decision module is used to fuse the basic reasoning results with the complex reasoning results, make a final decision based on the fusion results, and trigger the corresponding control operation according to the final decision.

[0043] Specifically, the fusion decision module fuses the basic reasoning result and the complex reasoning result by weighted average method. Specifically, the control operation triggered by the fusion decision module includes: adjusting attitude control and / or sending a flight alert.

[0044] Furthermore, in this embodiment, a data acquisition and preprocessing module is also provided in the microprocessor, and the data acquisition and preprocessing module is used to preprocess the collected data to meet the requirements of the neural network input in the NNabla Embedded module.

[0045] The steps for integrating the TensorFlow Lite for Microcontrollers module and the NNabla Embedded module are as follows:

[0046] (1) Data collection and preprocessing: Collect data from various sensors in the rocket core stage (such as accelerometers, gyroscopes, temperature sensors, and pressure sensors). Preprocess the collected data (such as normalization) to meet the requirements of neural network input.

[0047] (2) Basic reasoning (TensorFlow Lite for Microcontrollers): Input the preprocessed data into the TensorFlow Lite model for basic reasoning. The basic model reasoning is fast and suitable for real-time monitoring. Obtain the basic reasoning results and determine whether further complex reasoning is needed.

[0048] (3) Complex reasoning (NNabla Embedded): If the basic reasoning result exceeds the preset threshold, complex reasoning is triggered. The basic reasoning result and the original sensor data are input into the NNabla model for complex reasoning and optimization calculation. The complex reasoning result is obtained.

[0049] (4) Fusion decision: The basic reasoning results and the complex reasoning results are integrated using the weighted average method. The final decision is made based on the fusion results and the corresponding control operations are triggered (such as adjusting attitude control, sending flight alerts, etc.).

[0050] Specifically, in this embodiment, the software algorithm in the microprocessor adopts a layered architecture design, including:

[0051] Data acquisition layer: responsible for acquiring data from sensors and performing preliminary filtering and preprocessing;

[0052] Basic reasoning layer: Use TensorFlow Lite for Microcontrollers module for simple and efficient model reasoning;

[0053] Complex computing layer: Use NNabla Embedded modules to perform complex computing and optimization tasks;

[0054] Decision-making layer: Combines the results of basic reasoning and complex calculations to perform status monitoring and decision-making.

[0055] Specifically, in the embodiment of the present invention, the execution code in the microprocessor is as follows:

[0056]

[0057]

[0058] / / Define model and interpreter memory

[0059] static tflite::AllOpsResolver resolver;

[0060] static const tflite::Model*model=tflite::GetModel(your_tflite_model_data);

[0061] static uint8_t tensor_arena[10*1024];

[0062] static tflite::MicroInterpreter static_interpreter(

[0063] model,resolver,tensor_arena,sizeof(tensor_arena),nullptr);

[0064] tflite_interpreter=&static_interpreter;

[0065] / / Allocate tensor

[0066] tflite_interpreter->AllocateTensors();

[0067] }

[0068] / / Initialize NNablavoid setup_nnabla(){

[0069] nnabla_network.load_from_file("your_nnabla_model.nnb");

[0070] }

[0071] / / Data preprocessing

[0072] void preprocess_data(float*sensor_data,float*tflite_input){

[0073] / / Normalize the collected data to adapt it to the requirements of neural network input

[0074] for(int i=0; i <NUM_SENSORS;i++){

[0075] tflite_input[i] = sensor_data[i] / 1024.0f; / / Assume the sensor data range is 0 to 1024

[0076] }

[0077] }

[0078] / / Run TensorFlow Lite inference

[0079] void run_tflite_inference(float*input_data,float*output_data){

[0080] / / Get input and output tensors

[0081] float*input=tflite_interpreter->input(0)->data.f;

[0082] float*output=tflite_interpreter->output(0)->data.f;

[0083]

[0084]

[0085]

[0086] This embodiment provides a rocket core-stage flight status monitoring micromodule. The hot-swappable design allows the micromodule to be quickly plugged in and replaced, supporting rapid maintenance and fault replacement. In addition, the algorithm fusion design within the microprocessor integrates basic reasoning results and complex reasoning results, greatly reducing the performance requirements of the micromodule on the microprocessor. In addition, the micromodule has a universal interface, which allows the micromodule to be easily applied to different types of stand-alone boards, enhancing the scope of application and flexibility. : The modular design facilitates rapid replacement and upgrading, simplifies the maintenance process, and allows monitoring modules to be added or removed according to specific needs. The micro-modular design significantly reduces the size and weight of the system, helping to improve the payload efficiency and overall performance of the rocket.

[0087] Embodiment 2

[0088] Embodiment 2 of the present invention provides a design method for a rocket core-stage flight status monitoring micromodule, comprising the following steps:

[0089] Step 1: Determine the comprehensive cost target and volume and weight requirements of the module.

[0090] Specifically, the cost of a micromodule includes component cost, development cost, and production cost after design finalization and mass production. The actual comprehensive cost of a single micromodule is calculated based on the three.

[0091] Component cost (C e ): The sum of the costs of all components in a single micromodule.

[0092] Development cost (C d ): Fixed costs for designing, developing, and finalizing micromodules, including labor, equipment, and other R&D expenses.

[0093] Production cost (C u ): The variable cost of producing a single micromodule, including the costs of manufacturing, assembly, testing and quality control.

[0094] Production quantity (N): the number of micromodules produced in mass production.

[0095] In step 1, the calculation formula of the comprehensive cost target is:

[0096] C acc =C e +C d / N+C u ;

[0097] Among them, C acc represents the comprehensive cost, C e represents the development cost, C d represents the development cost, C u represents the mass production cost, and N represents the mass production quantity.

[0098] In addition, before conducting electrical design, the maximum volume and weight data envelope of the micromodule must be clarified based on the overall indicator requirements.

[0099] Step 2: Determine the module's power supply parameters, monitoring parameters, and environmental adaptation parameters.

[0100] This step is mainly about the design of universal requirements, which mainly include universal power supply, universal monitoring parameters, and universal environmental adaptability. The details are as follows:

[0101] Power supply: Make sure the power supply voltage input range is 3V to 3.6V, and the power supply is 3.3V under normal conditions;

[0102] Monitoring parameters: The number of monitoring channels of a single micromodule is not less than 8, the sampling rate of a single channel is not more than 32kHz, and the size is programmable;

[0103] Environmental adaptability: The module's operating temperature range is -40 to +85°C, ensuring the normal use of the rocket core stage under environmental conditions during mobile transportation and flight. The module's insulation resistance under normal conditions is not less than 100MΩ.

[0104] Step 3: Based on the comprehensive cost target, power supply parameters, monitoring parameters and environmental adaptability parameters, select and design the circuit components. When selecting, it is necessary to select the components based on comprehensive consideration of low cost, generalization and reliability.

[0105] Step 4: Select and design a connector suitable for hot plugging, and cooperate with the power input protection circuit and plug detection circuit to ensure that the microcontroller can sense the change of connection status during the hot plugging process.

[0106] Specifically, the power input protection circuit uses an N-channel MOSFET to control the connection and disconnection of the power supply. An RC delay circuit is added to the gate of the MOSFET to achieve a smooth rise in current and avoid surge current.

[0107] Specifically, the plug-in detection circuit uses a plug-in detection pin to monitor the plug-in status through a resistor voltage divider network and a comparator. When the module is detected to be plugged in or out, an interrupt signal is generated to notify the microcontroller.

[0108] Specifically, in the control circuit, the microcontroller reads the plug-in detection signal and controls the conduction of the MOSFET according to the state. At the same time, the voltage drop value on the current detection resistor is read through the AD core inside the microcontroller to monitor the system operating current during the micromodule plug-in process.

[0109] Step 5: Based on the selection of circuit components, design and draw the circuit schematic; at the same time, based on the volume and weight requirements of the module, perform PCB layout design.

[0110] Specifically, in this embodiment, an EDA design tool (Altium Designer) is used to draw a circuit schematic. When designing the PCB layout, the size and shape of the PCB are determined according to the volume and weight requirements in step 1, while optimizing the routing and device layout to reduce interference and signal attenuation.

[0111] Step 6: Architect the software algorithm for the microprocessor.

[0112] Compared with the best existing technology, the present invention has the following advantages:

[0113] Low cost: The manufacturing cost of the monitoring module is greatly reduced through optimized design and cost-effective device selection.

[0114] Universalization and compatibility: The universal interface is designed so that the micromodule can be easily applied to different types of stand-alone boards, enhancing the scope of application and flexibility.

[0115] Algorithm fusion design: realizes state monitoring.

[0116] Easy to maintain and expand: The modular design facilitates quick replacement and upgrade, simplifies the maintenance process, and allows monitoring modules to be added or removed according to specific needs.

[0117] Reduced Weight and Size: Micro-modular design significantly reduces the size and weight of the system, helping to improve the rocket's payload efficiency and overall performance.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rocket core-stage flight status monitoring micromodule, characterized in that: It includes a power supply, a sampling circuit, a hot-swap connector, a power input protection circuit, a plug-in detection circuit and a control circuit; The power supply supplies power to the hot-swappable connector through the power input protection circuit, the power input protection circuit uses an N-channel MOSFET to control the connection and disconnection of the power supply, and the gate of the MOSFET is provided with an RC delay circuit to achieve a smooth rise in current and avoid surge current; The plug-in detection circuit includes a resistor voltage divider network and a comparator, which is used to detect the plug-in status of the hot-plug connector and send it to the control circuit; The control circuit includes a microprocessor, which is used to read the plug-in detection signal to determine the plug-in state and control the MOSFET according to the plug-in state; it is also used to read the voltage drop value on the current detection resistor to monitor the system working current during the micromodule plug-in process; the microprocessor is provided with a TensorFlow Lite for Microcontrollers module and an NNablaEmbedded module; Among them, the TensorFlow Lite for Microcontrollers module is used to perform basic model reasoning according to the data input from the sensor, evaluate the basic state of the rocket core stage through basic model reasoning, and output the basic reasoning results of the state of the rocket core stage; the NNabla Embedded module is used to output the complex reasoning results of the state of the rocket core stage through the neural network model based on the basic reasoning results and the original sensor data when the basic reasoning results exceed the preset threshold.

2. A rocket core-stage flight status monitoring micromodule according to claim 1, characterized in that: The microprocessor is also provided with a fusion decision module, which is used to fuse the basic reasoning results with the complex reasoning results, make a final decision based on the fusion results, and trigger corresponding control operations according to the final decision.

3. A rocket core-stage flight status monitoring micromodule according to claim 2, characterized in that: The fusion decision module fuses the basic reasoning results with the complex reasoning results by weighted averaging.

4. A rocket core-stage flight status monitoring micromodule according to claim 2, characterized in that: The control operations triggered by the fusion decision module include: adjusting attitude control and / or sending a flight alert.

5. A rocket core-stage flight status monitoring micromodule according to claim 1, characterized in that: The microprocessor is also provided with a data acquisition and preprocessing module, and the data acquisition and preprocessing module is used to preprocess the collected data to meet the requirements of the neural network input in the NNabla Embedded module.

6. A rocket core-stage flight status monitoring micromodule according to claim 1, characterized in that: The software algorithm in the microprocessor is designed with a layered architecture, including: Data acquisition layer: responsible for acquiring data from sensors and performing preliminary filtering and preprocessing; Basic reasoning layer: Use TensorFlow Lite for Microcontrollers for simple and efficient model reasoning; Complex computing layer: Use NNabla Embedded to perform complex computing and optimization tasks; Decision-making layer: Combines the results of basic reasoning and complex calculations to perform status monitoring and decision-making.

7. The design method of a rocket core-stage flight status monitoring micromodule according to any one of claims 1 to 5, characterized in that: The following steps are involved: Step 1: Determine the comprehensive cost target, volume and weight requirements of the module; Step 2: Determine the power supply parameters, monitoring parameters and environmental adaptation parameters of the module; Step 3: Select and design circuit components based on comprehensive cost targets, power supply parameters, monitoring parameters, and environmental adaptability parameters; Step 4: Select and design a connector suitable for hot plugging, and cooperate with the power input protection circuit and plug detection circuit to ensure that the microcontroller can sense the change of connection status during the hot plugging process; Step 5: Based on the selection of circuit components, design and draw the circuit schematic; at the same time, based on the volume and weight requirements of the module, perform PCB layout design; Step 6: Architect the software algorithm for the microprocessor.

8. The design method of a rocket core-stage flight status monitoring micromodule according to claim 7 is characterized in that: In step 1, the calculation formula of the comprehensive cost target is: C acc =C e +C d / N+C u ; Among them, C acc represents the comprehensive cost, C e represents the development cost, C d represents the development cost, C u represents the mass production cost, and N represents the mass production quantity.

9. The design method of a rocket core-stage flight status monitoring micromodule according to claim 7 is characterized in that: In step 2, the parameters of the power supply include input voltage range and power supply voltage; Monitoring parameters include the number of monitoring channels and the single-channel sampling rate; environmental adaptation parameters include temperature range and insulation resistance value.

Images