CN117990111A - Method and system for planning partial path of lunar exploration robot based on MBSE model - Google Patents
Method and system for planning partial path of lunar exploration robot based on MBSE model Download PDFInfo
- Publication number
- CN117990111A CN117990111A CN202410397843.3A CN202410397843A CN117990111A CN 117990111 A CN117990111 A CN 117990111A CN 202410397843 A CN202410397843 A CN 202410397843A CN 117990111 A CN117990111 A CN 117990111A
- Authority
- CN
- China
- Prior art keywords
- data
- exploration robot
- lunar exploration
- planning
- path
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000011156 evaluation Methods 0.000 claims abstract description 35
- 238000004364 calculation method Methods 0.000 claims abstract description 20
- 238000003062 neural network model Methods 0.000 claims description 42
- 230000006870 function Effects 0.000 claims description 32
- 125000004122 cyclic group Chemical group 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000013500 data storage Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000012827 research and development Methods 0.000 abstract description 6
- 238000013528 artificial neural network Methods 0.000 abstract description 5
- 238000005457 optimization Methods 0.000 abstract description 5
- 239000000523 sample Substances 0.000 description 8
- 230000008569 process Effects 0.000 description 7
- 238000013461 design Methods 0.000 description 6
- 238000005265 energy consumption Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000000306 recurrent effect Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method and a system for planning a local path of a lunar exploration robot based on MBSE models, and relates to the technical field of lunar exploration path planning. Aiming at the problems of large calculation amount, high modeling cost and high difficulty in path planning and influence on the operation speed of the MBSE model in the modeling of the existing lunar exploration robot MBSE, the invention combines the historical data with the neural network, thereby not only fully utilizing the reference function of the historical data, but also being suitable for the conditions of various parameters, improving the efficiency of path planning and saving the research and development cost; aiming at the selection of sample data, an evaluation function is designed, high-quality parameters in historical data are reserved, and weight value setting in the evaluation function is carried out according to the relation between the nearest distance and the preset safety distance, so that the optimization of a prediction result can be ensured under different conditions.
Description
Technical Field
The invention relates to the technical field of lunar exploration path planning, in particular to a lunar exploration robot local path planning method and system based on MBSE models.
Background
MBSE (Model-Based SYSTEMS ENGINEERING), model-Based system engineering, for supporting system requirements, design, analysis, verification and validation activities, has the characteristics of integrating multidisciplinary environments, covering the entire life cycle, using standardized languages, automating environments, and combining with simulations, etc., and has been widely used in recent years.
In the modeling of the lunar exploration robot based on MBSE models, besides the conventional system design, complex and unknown lunar environments are more needed to be considered, and the lunar exploration robot is required to have autonomous navigation capability so as to effectively execute tasks. The path planning is a key technology for realizing autonomous navigation, and can help the lunar exploration robot to autonomously generate an optimal or suboptimal path from a starting point to a target point according to environmental information and task requirements.
The main methods for path planning of the lunar exploration robot in the prior art comprise an artificial potential field method, a genetic algorithm, a dynamic window method (DWA), a search-based method, an optimization-based method, a learning-based method and the like. However, these have advantages and disadvantages, and different planning methods are often required to be combined together to improve the effect and robustness of path planning, which results in reduced efficiency and increased calculation amount of path planning, further influences the running speed of MBSE models, and simultaneously has the problems of poor real-time performance, high energy consumption, low reliability, high modeling cost and high difficulty of path planning of the lunar exploration robot.
Therefore, when modeling the lunar exploration robot based on MBSE model, how to design a lunar exploration robot path planning method with simple algorithm and good effect is a direction which is still required to be continuously researched by the technicians in the field.
Disclosure of Invention
The invention provides a method and a system for planning a local path of a lunar exploration robot based on MBSE model, which are used for solving the problems of large path planning calculated amount, high modeling cost and high difficulty and influencing the operation speed of MBSE model in the modeling of the existing lunar exploration robot.
A lunar exploration robot local path planning method based on MBSE model specifically comprises the following steps:
Step S1, historical data of a lunar exploration robot local path plan is obtained, and the step comprises the following steps: raw data and planning data, the raw data comprising: the method comprises the steps of determining a speed parameter of a lunar exploration robot, a position parameter of the lunar exploration robot, a geometric parameter of the lunar exploration robot, a maximum speed parameter of the lunar exploration robot, a position parameter of an obstacle, a geometric parameter of the obstacle and a position parameter of a target point during path planning, wherein planning data comprise: the planned local path and the running parameters of the lunar exploration robot;
Step S2, obtaining data to be detected of a lunar exploration robot of a path to be planned, wherein the data comprise current speed parameters of the lunar exploration robot, current position parameters of the lunar exploration robot, geometric parameters of the lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters of an obstacle, geometric parameters of the obstacle and position parameters of a target point;
Step S3, calculating Euclidean distance between the data to be detected and each piece of original data in the historical data; if the minimum value of the Euclidean distance is smaller than or equal to a preset minimum value, planning data corresponding to the Euclidean distance are obtained, and the lunar exploration robot advances according to the planning data; if the Euclidean distance is greater than a preset minimum value, entering a step S4;
S4, inputting the data to be tested obtained in the step S2 into a trained neural network model, outputting planning data, and advancing the lunar exploration robot according to the planning data; the neural network model adopts a cyclic neural network model; the training method of the neural network model comprises the following steps:
step S41, evaluating the historical data obtained in the step S1, and taking the historical data meeting the evaluation requirement as sample data;
Step S42, training the neural network model through the sample data to obtain a trained neural network model; the input of the neural network model is the original data, and the output is the planning data;
And S5, storing the original data and planning data of the local path planning of the lunar exploration robot into historical data.
Further, in the step S3, the calculation formula of the euclidean distance is:
；
Wherein, O j is the Euclidean distance between the jth original data and the data to be measured, n is the number of parameter data in each original data, x ij is the ith parameter data in the jth original data, and x i0 is the ith parameter data in the data to be measured.
Further, in the step S4, the structure of the recurrent neural network model is as follows:
The input layer dimension is 7, the output layer dimension is 2, the number of hidden layer neurons is 24, the batch processing size is 32, the learning rate is 0.001, the iteration number is 100, and the optimizer is Adam.
Further, in the step S4, the loss function of the recurrent neural network model is a cross entropy loss function.
Further, in the step S41, the history data acquired in the step S1 is evaluated, and the adopted evaluation function is:
；
Wherein P is an evaluation value, R 1 is a value normalized by the current speed, ω 1 is a weight of the current speed, R 2 is a value normalized by the number of inflection points in the current path, ω 2 is a weight of the number of current inflection points, R 3 is a value normalized by the closest distance between the current path and the obstacle, ω 3 is a weight of the current closest distance, ω 1+ω2+ω3 =1.
Further, the determination method of omega 1 and omega 3 is as follows:
calculating the nearest distance between the lunar exploration robot and the obstacle according to the current position parameter of the lunar exploration robot, the geometric parameter of the lunar exploration robot, the position parameter of the obstacle and the geometric parameter of the obstacle, and determining omega 1 and omega 3 in an evaluation function according to the relationship between the nearest distance and a preset safety distance.
Further, the relationship between ω 1 and ω 3 is as follows:
Omega 1>ω3 when the nearest distance is greater than a preset safe distance, omega 1≤ω3 when the nearest distance is less than or equal to the preset safe distance.
Further, in the step S41, the evaluation requirement is: when the evaluation value P is greater than the preset evaluation value P 0, the requirement is satisfied.
A lunar exploration robot local path planning system based on MBSE model, the system adopts the lunar exploration robot local path planning method based on MBSE model as set forth in any one of the above, and specifically comprises the following modules:
A historical data acquisition module: the method for acquiring the historical data of the lunar exploration robot local path planning comprises the following steps: raw data and planning data, the raw data comprising: the method comprises the steps of determining a speed parameter of a lunar exploration robot, a position parameter of the lunar exploration robot, a geometric parameter of the lunar exploration robot, a maximum speed parameter of the lunar exploration robot, a position parameter of an obstacle, a geometric parameter of the obstacle and a position parameter of a target point during path planning, wherein planning data comprise: the planned local path and the running parameters of the lunar exploration robot;
the data acquisition module to be tested: the historical data acquisition module is connected with the historical data acquisition module and is used for acquiring current data of the lunar exploration robot of the path to be planned, wherein the current data comprise current speed parameters of the lunar exploration robot, current position parameters of the lunar exploration robot, geometric parameters of the lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters of the obstacle, geometric parameters of the obstacle and position parameters of the target point;
The Euclidean distance calculation module: the data acquisition module is connected with the data acquisition module and used for calculating the Euclidean distance between the data to be detected and each piece of original data in the historical data; if the minimum value of the Euclidean distance is smaller than or equal to a preset minimum value, planning data corresponding to the Euclidean distance are obtained, and the lunar exploration robot advances according to the planning data; if the Euclidean distance is greater than a preset minimum value, entering a path prediction module;
And a path prediction module: the system comprises a Euclidean distance calculation module, a neural network model, a planning data acquisition module, a lunar exploration robot and a lunar exploration robot, wherein the Euclidean distance calculation module is connected with the Euclidean distance calculation module and is used for inputting data to be measured into the trained neural network model and outputting the planning data, and the lunar exploration robot advances according to the planning data; the neural network model adopts a cyclic neural network model;
And a data storage module: and the path prediction module is connected with the path prediction module and is used for storing the original data and the planning data of the local path planning of the current lunar exploration robot into the historical data.
Compared with the prior art, the invention has the beneficial effects that:
Firstly, the invention combines the historical data with the neural network, thereby fully utilizing the reference function of the historical data, being applicable to the conditions of various parameters, reducing the calculated amount of path planning, improving the running speed of MBSE model, improving the efficiency of path planning and saving the research and development cost;
Secondly, the Euclidean distance between the data to be detected and the historical data is calculated, and when the Euclidean distance is smaller than a preset value, the lunar exploration robot advances according to the path parameters and the motion parameters corresponding to the historical data, so that the reference function of the historical data is fully utilized, and the path planning efficiency is improved;
thirdly, an evaluation function is designed during sample data screening, historical data is screened according to a calculation result, and more high-quality historical data is reserved, so that a path and motion parameters of a neural network model planning are more accurate, and stability and robustness are improved;
Fourth, when the evaluation function is designed, important parameters of path planning of the lunar exploration robot, including the running speed, the safety distance and the quantity of inflection points, are fully considered, so that efficient execution of tasks and safety of the robot are ensured;
Fifthly, when the motion parameters of the planned local path and the lunar exploration robot are evaluated, the weight value is set according to the relation between the nearest distance and the preset safety distance, so that the optimization of the prediction result can be ensured under different conditions.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for planning a local path of a lunar exploration robot based on MBSE models;
Fig. 2 is a schematic structural diagram of a lunar exploration robot local path planning system based on MBSE models.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the invention, are within the scope of the invention.
The following describes specific embodiments of the present invention with reference to the drawings.
The MBSE model is used for supporting system requirements, design, analysis, verification and verification activities, has the characteristics of integrating multidisciplinary environments, covering the whole life cycle, using standardized languages, automating environments, combining with simulation and the like, and has been widely applied in recent years. The system modeling is carried out on the lunar exploration robot through the MBSE model, the design scheme is visual and visual, the system structure and the function are clear and definite, the design quality and the development efficiency can be improved, and the research and development cost is saved. Modeling a lunar exploration robot based on MBSE models generally includes the steps of:
Step SA1, determining a functional index and a nonfunctional index of the lunar exploration robot; the functional index comprises: mobile function requirements, survey function requirements, transfer function requirements, transport function requirements, retrofit function requirements, and excavation function requirements; the non-functional index includes: weight demand, power demand, size demand, probe distance demand, speed demand, obstacle crossing demand, maximum output force demand, and maximum working distance demand;
Step SA2, inputting the functional index and the nonfunctional index into SysML modeling software to generate a demand graph; accurately expressing the ambiguity requirement in the requirement graph to obtain a technical index;
step SA3, determining the system composition of the lunar exploration robot according to the technical indexes; the system comprises a control system, a mechanical arm operation system, a mobile system, an energy system, a communication system and a refitting system;
step SA4, determining the internal composition relation and interface relation of each system;
Step SA5, constructing a lunar exploration robot MBSE model by adopting MAGIC DRAW software according to the system architecture model, the internal composition relation and the interface relation of each system;
Step SA6, verifying the lunar exploration robot model by adopting simulation software, and specifically comprises the steps of establishing a SysML model by utilizing MAGIC DRAW software, and verifying the functional index and the nonfunctional index of a system architecture; establishing a three-dimensional solid model of the robot by using CATIA software, and verifying the three-dimensional solid model; establishing a mechanical arm model by MWorks software, and verifying the mechanical arm model;
and step SA7, optimizing the lunar exploration robot model according to the verification result.
In the modeling process, if the calculation amount of path planning is large, the modeling cost is high, the difficulty is large, and meanwhile, the running speed of the MBSE model is also reduced. The invention combines the historical data with the neural network, thereby fully utilizing the reference function of the historical data, being applicable to the conditions of various parameters, improving the efficiency of path planning, reducing the calculated amount of path planning, improving the running speed of MBSE models and saving the research and development cost; aiming at the selection of sample data, an evaluation function is designed, high-quality parameters in historical data are reserved, and weight value setting in the evaluation function is carried out according to the relation between the nearest distance and the preset safety distance, so that the optimization of a prediction result can be ensured under different conditions.
Example 1
As shown in fig. 1, the invention provides a lunar exploration robot local path planning method based on MBSE model, which specifically comprises the following steps:
step S1, historical data of a lunar exploration robot local path plan is obtained, and the step comprises the following steps: raw data and planning data, the raw data comprising: the method comprises the steps of determining a speed parameter of a lunar exploration robot, a position parameter of the lunar exploration robot, a geometric parameter of the lunar exploration robot, a maximum speed parameter of the lunar exploration robot, a position parameter of an obstacle, a geometric parameter of the obstacle and a position parameter of a target point during path planning, wherein planning data comprise: and planning a local path and operating parameters of the lunar exploration robot.
Step S2, obtaining data to be detected of a lunar exploration robot of a path to be planned, wherein the data comprise current speed parameters of the lunar exploration robot, current position parameters of the lunar exploration robot, geometric parameters of the lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters of an obstacle, geometric parameters of the obstacle and position parameters of a target point.
Step S3, calculating Euclidean distance between the data to be detected and each piece of original data in the historical data; if the minimum value of the Euclidean distance is smaller than or equal to a preset minimum value, planning data corresponding to the Euclidean distance are obtained, and the lunar exploration robot advances according to the planning data; if the Euclidean distance is greater than a preset minimum value, entering a step S4;
in the step S3, the calculation formula of the euclidean distance is as follows:
；
Wherein, O j is the Euclidean distance between the jth original data and the data to be measured, n is the number of parameter data in each original data, x ij is the ith parameter data in the jth original data, and x i0 is the ith parameter data in the data to be measured.
Euclidean distance is a measure of signal similarity. In signal processing, the smaller the Euclidean distance of two signals, the more similar they are; conversely, the larger the Euclidean distance, the larger their difference. The invention searches the data closest to the data to be detected in the historical data by calculating the Euclidean distance. When the Euclidean distance is smaller than a preset value, the lunar exploration robot advances according to the path parameters and the motion parameters corresponding to the historical data, the reference function of the historical data is fully utilized, the path planning efficiency is improved, and the research and development cost is saved.
S4, inputting the data to be tested obtained in the step S2 into a trained neural network model, outputting planning data, and advancing the lunar exploration robot according to the planning data; the neural network model adopts a cyclic neural network model;
when the lunar exploration robot is used for planning a local path, input data comprise the current speed parameter and the position parameter of the lunar exploration robot, the geometric parameter of the lunar exploration robot, the maximum speed parameter of the lunar exploration robot, the position parameter and the geometric parameter of an obstacle, and the sequence output of the position parameter of a target point, wherein the position data and the speed data have a corresponding relation in time. Therefore, the neural network model which can benefit the sequence data is selected, and the high efficiency and accuracy of the prediction result are improved.
The cyclic neural network model (RNN) is particularly suitable for processing sequence data, and the dynamic nature of the RNN also enables it to adapt to dynamically changing lunar environments, in addition to which smooth trajectories can be generated by the RNN, reducing energy consumption and mechanical wear, while improving the movement efficiency of the robot. This feature makes the RNN suitable for local path planning tasks that require real-time response.
The structure of the cyclic neural network model adopted in the invention is as follows:
The input layer dimension is 7, the output layer dimension is 2, the hidden layer neuron number is 24, the batch processing size is 32, the learning rate is 0.001, the iteration number is 100, and the optimizer is Adam.
The 7 dimensions of the input layer correspond to the 7 parameters input respectively: the method comprises the steps of current speed parameters, position parameters, geometric parameters of a lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters and geometric parameters of an obstacle and position parameters of a target point of the lunar exploration robot;
The 7 dimensions of the output layer correspond to the output parameters in 2: parameters of planned path and lunar exploration robot motion.
The loss function of the recurrent neural network model is a cross entropy loss function.
In a sequence generation task, the output of each time step may be a class, so cross entropy loss may be used to process such tasks. At each time step, the RNN output is compared to the true next marker and the cross entropy loss is calculated. The cross entropy loss function has the advantages that the difference between model prediction and real distribution can be effectively measured, the model can be promoted to be quickly converged, and the accuracy of classification problems can be improved.
The training method of the neural network model comprises the following steps:
step S41, evaluating the historical data obtained in the step S1, and taking the historical data meeting the evaluation requirement as sample data;
Step S42, training the neural network model through the sample data to obtain a trained neural network model; and the input of the neural network model is the original data, and the output is the planning data.
In the step S41, the history data acquired in the step S1 is evaluated, and the adopted evaluation function is:
；
Wherein P is an evaluation value, R 1 is a value normalized by the current speed, ω 1 is a weight of the current speed, R 2 is a value normalized by the number of inflection points in the current path, ω 2 is a weight of the number of current inflection points, R 3 is a value normalized by the closest distance between the current path and the obstacle, ω 3 is a weight of the current closest distance, ω 1+ω2+ω3 =1.
Further, the determination method of omega 1 and omega 3 is as follows:
calculating the nearest distance between the lunar exploration robot and the obstacle according to the current position parameter of the lunar exploration robot, the geometric parameter of the lunar exploration robot, the position parameter of the obstacle and the geometric parameter of the obstacle, and determining omega 1 and omega 3 in an evaluation function according to the relationship between the nearest distance and a preset safety distance.
Further, the relationship between ω 1 and ω 3 is as follows:
Omega 1>ω3 when the nearest distance is greater than a preset safe distance, omega 1≤ω3 when the nearest distance is less than or equal to the preset safe distance.
When the lunar exploration robot performs path planning, multiple factors such as running speed, safety distance, smoothness and the like must be comprehensively considered so as to ensure efficient execution of tasks and safety of the robot;
Consider the running speed of the lunar exploration robot: the running speed of the robot directly affects the time efficiency of task completion. In the path planning process, the optimal running speed needs to be determined according to the task requirements, the terrain complexity and the mechanical performance of the robot. Too fast a speed may cause the robot to lose stability or not accurately sense and respond to environmental changes, while too slow a speed may lengthen task execution time and increase energy consumption.
Consider the safety distance of the lunar exploration robot: safety distance refers to the minimum distance a robot maintains from an obstacle or other source of danger during movement. In path planning, the size, shape and position of the obstacle and the obstacle avoidance capability of the robot must be fully considered, so that the robot can maintain a sufficient safety distance in the whole moving process. This helps to reduce the risk of collision of the robot with obstacles, protecting the safety of the robot.
Consider the smoothness of the running path of the lunar exploration robot: the smoothness of the path has an important influence on the movement performance and the energy consumption of the robot. In path planning, abrupt actions such as sharp turning, sudden braking and the like should be avoided as much as possible so as to reduce mechanical abrasion and energy consumption of the robot. By optimizing the smoothness of the path, the movement of the robot can be smoother and more stable, and the efficiency and reliability of task execution are improved.
In a specific embodiment, when the nearest distance is greater than a preset safety distance, ω 1=0.5,ω3=0.3,ω2 =0.2, and when the nearest distance is less than or equal to the preset safety distance, ω 1=0.3,ω3=0.5,ω2 =0.2, and compared with ω 1=ω2=ω3, the efficiency is improved by 8% in the same path.
The evaluation function designed by the invention considers the speed, the quantity of inflection points and the nearest distance, is used for evaluating the path and the motion parameters output by the neural network, and considers the motion speed to improve the efficiency when the robot is far away from the obstacle, and considers the safety when the robot is near to the obstacle, and considers the smoothness of the path when the quantity of inflection points is the same; when the planned local path and the motion parameters of the lunar exploration robot are evaluated, weight value setting is carried out according to the relationship between the nearest distance and the preset safety distance, so that the optimization of the prediction result can be ensured under different conditions.
The requirements of the evaluation result are: when the evaluation value P is greater than the preset evaluation value P 0, the requirement is satisfied. The evaluation value is generally used for measuring the performance of a model or an algorithm, experiments are carried out in actual data or a simulation environment, the rationality and the effectiveness of the evaluation value are verified, and when the evaluation value is set, the limitation and the requirement of actual application, such as physical characteristics of a robot, uncertainty of the environment and the like, need to be considered. P 0 is adjustable according to the actual test condition. In one embodiment, the preset evaluation value P 0 is set to 80%, and the path feasibility is significantly improved compared to the case where no evaluation is performed.
According to the invention, during sample data screening, an evaluation function is designed, historical data is screened according to a calculation result, and more high-quality historical data is reserved, so that the path and motion parameters planned by the neural network model are more accurate, and the stability and robustness are improved; when the evaluation function is designed, important parameters of path planning of the lunar exploration robot, including the running speed, the safety distance and the quantity of inflection points, are fully considered, and efficient execution of tasks and safety of the robot are ensured.
And S5, storing the original data and planning data of the local path planning of the lunar exploration robot into historical data.
The invention combines the historical data with the neural network, thereby not only fully utilizing the reference function of the historical data, but also being suitable for the conditions of various parameters, reducing the calculated amount of path planning, improving the running speed of MBSE model, improving the efficiency of path planning and saving the research and development cost.
Example 2
As shown in fig. 2, the invention further provides a system for planning a local path of a lunar exploration robot based on MBSE model, wherein the system adopts the method for planning a local path of a lunar exploration robot based on MBSE model according to any one of embodiment 1, and specifically comprises the following modules:
A historical data acquisition module: the method for acquiring the historical data of the lunar exploration robot local path planning comprises the following steps: raw data and planning data, the raw data comprising: the method comprises the steps of determining a speed parameter of a lunar exploration robot, a position parameter of the lunar exploration robot, a geometric parameter of the lunar exploration robot, a maximum speed parameter of the lunar exploration robot, a position parameter of an obstacle, a geometric parameter of the obstacle and a position parameter of a target point during path planning, wherein planning data comprise: the planned local path and the running parameters of the lunar exploration robot;
the data acquisition module to be tested: the historical data acquisition module is connected with the historical data acquisition module and is used for acquiring current data of the lunar exploration robot of the path to be planned, wherein the current data comprise current speed parameters of the lunar exploration robot, current position parameters of the lunar exploration robot, geometric parameters of the lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters of the obstacle, geometric parameters of the obstacle and position parameters of the target point;
The Euclidean distance calculation module: the data acquisition module is connected with the data acquisition module and used for calculating the Euclidean distance between the data to be detected and each piece of original data in the historical data; if the minimum value of the Euclidean distance is smaller than or equal to a preset minimum value, planning data corresponding to the Euclidean distance are obtained, and the lunar exploration robot advances according to the planning data; if the Euclidean distance is greater than a preset minimum value, entering a path prediction module;
And a path prediction module: the system comprises a Euclidean distance calculation module, a neural network model, a planning data acquisition module, a lunar exploration robot and a lunar exploration robot, wherein the Euclidean distance calculation module is connected with the Euclidean distance calculation module and is used for inputting data to be measured into the trained neural network model and outputting the planning data, and the lunar exploration robot advances according to the planning data; the neural network model adopts a cyclic neural network model;
And a data storage module: and the path prediction module is connected with the path prediction module and is used for storing the original data and the planning data of the local path planning of the current lunar exploration robot into the historical data.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present invention. As used in this specification, the terms "a," "an," "the," and/or "the" are not intended to be limiting, but rather are to be construed as covering the singular and the plural, unless the context clearly dictates otherwise. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method or apparatus that includes the element.
It should also be noted that the positional or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the positional or positional relationship shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or element in question must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. Unless specifically stated or limited otherwise, the terms "mounted," "connected," and the like are to be construed broadly and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.
Claims (9)
1. A lunar exploration robot local path planning method based on MBSE model is characterized by comprising the following steps:
Step S1, historical data of a lunar exploration robot local path plan is obtained, and the step comprises the following steps: raw data and planning data, the raw data comprising: the method comprises the steps of determining a speed parameter of a lunar exploration robot, a position parameter of the lunar exploration robot, a geometric parameter of the lunar exploration robot, a maximum speed parameter of the lunar exploration robot, a position parameter of an obstacle, a geometric parameter of the obstacle and a position parameter of a target point during path planning, wherein planning data comprise: the planned local path and the running parameters of the lunar exploration robot;
Step S2, obtaining data to be detected of a lunar exploration robot of a path to be planned, wherein the data comprise current speed parameters of the lunar exploration robot, current position parameters of the lunar exploration robot, geometric parameters of the lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters of an obstacle, geometric parameters of the obstacle and position parameters of a target point;
Step S3, calculating Euclidean distance between the data to be detected and each piece of original data in the historical data; if the minimum value of the Euclidean distance is smaller than or equal to a preset minimum value, planning data corresponding to the Euclidean distance are obtained, and the lunar exploration robot advances according to the planning data; if the Euclidean distance is greater than a preset minimum value, entering a step S4;
S4, inputting the data to be tested obtained in the step S2 into a trained neural network model, outputting planning data, and advancing the lunar exploration robot according to the planning data; the neural network model adopts a cyclic neural network model, and the training method comprises the following steps:
step S41, evaluating the historical data obtained in the step S1, and taking the historical data meeting the evaluation requirement as sample data;
Step S42, training the neural network model through the sample data to obtain a trained neural network model; the input of the neural network model is the original data, and the output is the planning data;
And S5, storing the original data and planning data of the local path planning of the lunar exploration robot into historical data.
2. The method for planning a local path of a lunar exploration robot based on MBSE model according to claim 1, wherein the calculation formula of the euclidean distance in step S3 is:
；
Wherein, O j is the Euclidean distance between the jth original data and the data to be measured, n is the number of parameter data in each original data, x ij is the ith parameter data in the jth original data, and x i0 is the ith parameter data in the data to be measured.
3. The method for planning a local path of a lunar exploration robot based on MBSE model according to claim 1, wherein in step S4, the neural network model has the structure that:
The input layer dimension is 7, the output layer dimension is 2, the number of hidden layer neurons is 24, the batch processing size is 32, the learning rate is 0.001, the iteration number is 100, and the optimizer is Adam.
4. The method for planning a local path of a lunar exploration robot based on MBSE model according to claim 1, wherein in step S4, the loss function of the neural network model is a cross entropy loss function.
5. The method for planning a local path of a lunar exploration robot based on MBSE model according to claim 1, wherein in step S41, an evaluation function is adopted when evaluating the historical data acquired in step S1, where:
；
Wherein P is an evaluation value, R 1 is a value normalized by the current speed, ω 1 is a weight of the current speed, R 2 is a value normalized by the number of inflection points in the current path, ω 2 is a weight of the number of current inflection points, R 3 is a value normalized by the closest distance between the current path and the obstacle, ω 3 is a weight of the current closest distance, ω 1+ω2+ω3 =1.
6. The method for planning a local path of a lunar exploration robot based on MBSE model according to claim 5, wherein the determining method of ω 1 and ω 3 is:
calculating the nearest distance between the lunar exploration robot and the obstacle according to the current position parameter of the lunar exploration robot, the geometric parameter of the lunar exploration robot, the position parameter of the obstacle and the geometric parameter of the obstacle, and determining omega 1 and omega 3 in an evaluation function according to the relationship between the nearest distance and a preset safety distance.
7. The method for planning a local path of a lunar exploration robot based on MBSE model of claim 6, wherein the relationship between ω 1 and ω 3 is as follows:
Omega 1>ω3 when the nearest distance is greater than a preset safe distance, omega 1≤ω3 when the nearest distance is less than or equal to the preset safe distance.
8. The method for planning a local path of a lunar exploration robot based on MBSE model according to claim 5, wherein in step S41, the evaluation requirement is: when the evaluation value P is greater than the preset evaluation value P 0, the requirement is satisfied.
9. A lunar exploration robot local path planning system based on MBSE model, which is characterized in that the system adopts the lunar exploration robot local path planning method based on MBSE model as set forth in any one of claims 1 to 8, and specifically comprises the following modules:
A historical data acquisition module: the method for acquiring the historical data of the lunar exploration robot local path planning comprises the following steps: raw data and planning data, the raw data comprising: the method comprises the steps of determining a speed parameter of a lunar exploration robot, a position parameter of the lunar exploration robot, a geometric parameter of the lunar exploration robot, a maximum speed parameter of the lunar exploration robot, a position parameter of an obstacle, a geometric parameter of the obstacle and a position parameter of a target point during path planning, wherein planning data comprise: the planned local path and the running parameters of the lunar exploration robot;
the data acquisition module to be tested: the historical data acquisition module is connected with the historical data acquisition module and is used for acquiring current data of the lunar exploration robot of the path to be planned, wherein the current data comprise current speed parameters of the lunar exploration robot, current position parameters of the lunar exploration robot, geometric parameters of the lunar exploration robot, maximum speed parameters of the lunar exploration robot, position parameters of the obstacle, geometric parameters of the obstacle and position parameters of the target point;
The Euclidean distance calculation module: the data acquisition module is connected with the data acquisition module and used for calculating the Euclidean distance between the data to be detected and each piece of original data in the historical data; if the minimum value of the Euclidean distance is smaller than or equal to a preset minimum value, planning data corresponding to the Euclidean distance are obtained, and the lunar exploration robot advances according to the planning data; if the Euclidean distance is greater than a preset minimum value, entering a path prediction module;
And a path prediction module: the system comprises a Euclidean distance calculation module, a neural network model, a planning data acquisition module, a lunar exploration robot and a lunar exploration robot, wherein the Euclidean distance calculation module is connected with the Euclidean distance calculation module and is used for inputting data to be measured into the trained neural network model and outputting the planning data, and the lunar exploration robot advances according to the planning data; the neural network model adopts a cyclic neural network model;
And a data storage module: and the path prediction module is connected with the path prediction module and is used for storing the original data and the planning data of the local path planning of the current lunar exploration robot into the historical data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410397843.3A CN117990111B (en) | 2024-04-03 | 2024-04-03 | Method and system for planning partial path of lunar exploration robot based on MBSE model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410397843.3A CN117990111B (en) | 2024-04-03 | 2024-04-03 | Method and system for planning partial path of lunar exploration robot based on MBSE model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117990111A true CN117990111A (en) | 2024-05-07 |
| CN117990111B CN117990111B (en) | 2024-07-12 |
Family
ID=90895462
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410397843.3A Active CN117990111B (en) | 2024-04-03 | 2024-04-03 | Method and system for planning partial path of lunar exploration robot based on MBSE model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117990111B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109765626A (en) * | 2019-02-20 | 2019-05-17 | 吉林大学 | A kind of lunar exploration radar data processing method based on the offset of least square kirchhoff |
| CN111580519A (en) * | 2020-05-12 | 2020-08-25 | 北京控制工程研究所 | Quasi-real-time teleoperation system for lunar surface detection under earth-moon time delay |
| JP2021034050A (en) * | 2019-08-21 | 2021-03-01 | 哈爾浜工程大学 | Auv action plan and operation control method based on reinforcement learning |
| CN115540878A (en) * | 2022-09-27 | 2022-12-30 | 北京航天飞行控制中心 | Lunar surface driving navigation method and device, electronic equipment and storage medium |
| CN115933652A (en) * | 2022-11-29 | 2023-04-07 | 北京航天飞行控制中心 | Lunar vehicle direct-drive teleoperation driving method based on sequence image splicing and fusion |
| CN117308968A (en) * | 2023-08-31 | 2023-12-29 | 上海宇航系统工程研究所 | Moon surface large-range path planning system and method of BS architecture |
| CN117516572A (en) * | 2023-11-20 | 2024-02-06 | 中国人民解放军战略支援部队航天工程大学 | Lunar surface path planning method based on reinforcement learning |
| CN117707168A (en) * | 2023-12-18 | 2024-03-15 | 四川华控图形科技有限公司 | Robot obstacle avoidance path planning method based on deep reinforcement learning |
-
2024
- 2024-04-03 CN CN202410397843.3A patent/CN117990111B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109765626A (en) * | 2019-02-20 | 2019-05-17 | 吉林大学 | A kind of lunar exploration radar data processing method based on the offset of least square kirchhoff |
| JP2021034050A (en) * | 2019-08-21 | 2021-03-01 | 哈爾浜工程大学 | Auv action plan and operation control method based on reinforcement learning |
| CN111580519A (en) * | 2020-05-12 | 2020-08-25 | 北京控制工程研究所 | Quasi-real-time teleoperation system for lunar surface detection under earth-moon time delay |
| CN115540878A (en) * | 2022-09-27 | 2022-12-30 | 北京航天飞行控制中心 | Lunar surface driving navigation method and device, electronic equipment and storage medium |
| CN115933652A (en) * | 2022-11-29 | 2023-04-07 | 北京航天飞行控制中心 | Lunar vehicle direct-drive teleoperation driving method based on sequence image splicing and fusion |
| CN117308968A (en) * | 2023-08-31 | 2023-12-29 | 上海宇航系统工程研究所 | Moon surface large-range path planning system and method of BS architecture |
| CN117516572A (en) * | 2023-11-20 | 2024-02-06 | 中国人民解放军战略支援部队航天工程大学 | Lunar surface path planning method based on reinforcement learning |
| CN117707168A (en) * | 2023-12-18 | 2024-03-15 | 四川华控图形科技有限公司 | Robot obstacle avoidance path planning method based on deep reinforcement learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117990111B (en) | 2024-07-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102292277B1 (en) | LIDAR localization inferring solutions using 3D CNN networks in autonomous vehicles | |
| CN112356830B (en) | Intelligent parking method based on model reinforcement learning | |
| CN109755995B (en) | Robot automatic charging docking method based on ROS robot operating system | |
| KR102350181B1 (en) | LIDAR Position Estimation Using RNN and LSTM to Perform Temporal Smoothing in Autonomous Vehicles | |
| CN111695299B (en) | Mesoscale vortex trajectory prediction method | |
| CN111971574A (en) | Deep learning based feature extraction for LIDAR localization of autonomous vehicles | |
| Wen et al. | MRPB 1.0: A unified benchmark for the evaluation of mobile robot local planning approaches | |
| CN108958238B (en) | Robot point-to-area path planning method based on covariant cost function | |
| Tsardoulias et al. | Critical rays scan match SLAM | |
| Saulnier et al. | Information theoretic active exploration in signed distance fields | |
| Zhang et al. | Multiple model AUV navigation methodology with adaptivity and robustness | |
| CN111190211B (en) | GPS failure position prediction positioning method | |
| CN110895146A (en) | Synchronous positioning and map construction method for mobile robot | |
| Luo et al. | A multi-scale map method based on bioinspired neural network algorithm for robot path planning | |
| CN117141520B (en) | Real-time track planning method, device and equipment | |
| CN116337045A (en) | High-speed map building navigation method based on karto and teb | |
| CN114237256A (en) | Three-dimensional path planning and navigation method suitable for under-actuated robot | |
| CN115933641A (en) | AGV path planning method based on model prediction control guidance deep reinforcement learning | |
| Guo et al. | Local path planning of mobile robot based on long short-term memory neural network | |
| CN114386599B (en) | Method and device for training trajectory prediction model and trajectory planning | |
| Zhang et al. | Neural network based uncertainty prediction for autonomous vehicle application | |
| Cai et al. | A prior information‐based coverage path planner for underwater search and rescue using autonomous underwater vehicle (AUV) with side‐scan sonar | |
| Cen | Target location detection of mobile robots based on R-FCN deep convolutional neural network | |
| CN117990111B (en) | Method and system for planning partial path of lunar exploration robot based on MBSE model | |
| CN116907510A (en) | Intelligent motion recognition method based on Internet of things technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |