Boat racing simulation training method and system and storage medium

Abstract

The invention discloses a racing boat simulation training method and system and a storage medium, and the method comprises the steps: obtaining the light emitted by a light source of a water surface where an unmanned boat is located, the reflected light of the surface of a water surface object, and the water bottom refracted light, constructing an optical model of a water surface environment, and obtaining the pose information returned by the unmanned boat when the unmanned boat sails on the water surface; constructing an unmanned ship motion model according to the pose information and the optical model, performing dynamic simulation on the racing ship according to the unmanned ship motion model to determine a dynamic simulation model of the racing ship, acquiring training data of the racing ship, and preprocessing the training data to obtain a racing ship association rule; based on the racing boat association rule and the dynamics simulation model, a regression analysis algorithm is adopted to carry out racing boat score prediction so as to complete racing boat simulation training, a more vivid simulation scene of racing boat can be effectively provided, and a training plan and training score prediction can be automatically generated by analyzing the racing boat association rule. The problems of excessive training or insufficient training and the like of athletes are effectively avoided, and the training level is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of simulation training, and in particular relates to a rowing simulation training method, system and storage medium. Background Art

[0002] Rowing is a water sport that uses oars as a propulsion force to propel a boat forward on the water. Rowing usually involves fluid mechanics, materials science, etc. From the perspective of fluid mechanics, the shape and blades of a rowing boat must have the shape and performance with the least resistance; in terms of materials, a rowing boat must have certain strength, durability and other properties, while being lightweight. However, rowing competitions are mostly held on calm waters, but the water surface during actual rowing training is mostly not calm, which makes rowing training limited and the training data provided to athletes is not very accurate, thus affecting the effect of rowing training. Summary of the invention

[0003] In view of this, the present invention provides a rowing simulation training method, system and storage medium that can collect rowing training data in real time and improve rowing training effects, so as to solve the above-mentioned technical problems, and specifically adopts the following technical solutions to achieve this.

[0004] In a first aspect, the present invention provides a rowing simulation training method, comprising the following steps:

[0005] Obtain the light emitted by the light source on the water surface where the unmanned boat is located, the reflected light from the surface of the water object, and the refracted light from the bottom of the water, and build an optical model of the water surface environment;

[0006] Obtain the posture information sent back by the unmanned boat while sailing on the water surface, and build the unmanned boat motion model based on the posture information and the optical model;

[0007] Perform dynamic simulation on the racing boat according to the unmanned boat motion model to determine the dynamic simulation model of the racing boat;

[0008] Acquire historical rowing training data and pre-process the training data to obtain rowing association rules, wherein the historical training data includes data records of rowing athletes' daily training and testing, and rowing training time series data;

[0009] Based on the rowing association rules and dynamic simulation model, a regression analysis algorithm is used to predict rowing results to complete rowing simulation training.

[0010] As a preferred embodiment of the above technical solution, the position and posture information transmitted by the unmanned boat while sailing on the water surface is obtained, and the motion model of the unmanned boat is constructed according to the position and posture information and the optical model, including:

[0011] Initialize the parameters of the unmanned boat and perform cyclic calculations based on the input propeller angle and engine speed. The cyclic calculations include calculating the fluid dynamics and torque, propeller thrust and torque, and wind and wave interference forces and torques.

[0012] Construct the inertial coordinate system and attached coordinate system of the unmanned boat when it moves on the water surface, and divide the displacement and rotation of the unmanned boat into three degrees of freedom;

[0013] According to the corresponding relationship between the physical quantities of the inertial coordinate system and the attached coordinate system, two sets of expressions (1) and (2) are obtained;

[0014]

[0015] Among them, the physical quantities belonging to the inertial coordinate system are: the forces X0, Y0 and moment N0 acting on the boat, the coordinates x of the center of gravity G of the unmanned boat G0 ,y G0 ; The attached coordinate system includes: the forces X, Y on the hull, and the external torque. The x, Y, V, and r represent the forward speed, drift speed, and bow angular velocity of the unmanned boat, respectively. c Indicates the x-axis coordinate value of the center of the unmanned boat;

[0016] As a preferred embodiment of the above technical solution, the unmanned boat in the preset inertial coordinate system is a rigid body. During the force movement, its shape, size and relative positions of different points in the boat body do not change. According to Newton's second law:

[0017]

[0018] Among them, m and I zz They represent the mass of the unmanned boat and the mass moment of inertia of the unmanned boat around the x-axis, I zz Expressed as L, B, C b Respectively represent the length, width, average draft and square coefficient of the unmanned boat, x' G0 ,y' G0 They represent the linear acceleration of the unmanned boat along the x0 and y0 directions respectively, and ψ' represents the bow acceleration of the unmanned boat;

[0019] After taking derivatives of the variables in expression (2), the variables are introduced into the basic motion equation of expression (3) and projected into the attached coordinate system. After calculation, the relational expression of the unmanned boat motion model is derived as follows:

[0020]

[0021] Among them, u', v', and r' respectively represent the forward speed, drift speed, and bow angular velocity in the unmanned boat motion model. Expanding expression (4) to obtain the complete unmanned boat motion model relationship expression is:

[0022]

[0023] Among them, the subscript H0 represents the fluid dynamic force and torque, the subscript P represents the propeller thrust and torque, the subscript WIND and WAVE represent the force and torque generated by wind and waves, and the correction term Y·x c Credit middle.

[0024] As a preferred embodiment of the above technical solution, the fluid power on the unmanned boat is the reaction force generated by the fluid moving in the fluid on the surface of the boat body in contact with the unmanned boat, and the reaction force includes inertial force and viscous force. This breaks down into:

[0025]

[0026] Among them, the subscript I represents inertial force, and H represents viscous force;

[0027] The thrust vector model is used to divide the force generated by the propeller into thrust T and lateral pulling force N on the unmanned boat during steering. The preset propeller turning angle is δ, and the longitudinal and lateral thrusts of the propeller on the unmanned boat are:

[0028]

[0029] Among them, t P represents the thrust derating factor, a P represents the lateral force coefficient, X P Indicates the Y-axis offset of the center of the thruster in the attached coordinate system.

[0030] As a preferred embodiment of the above technical solution, rowing training data is obtained and the rowing association rules are obtained by preprocessing the training data, including:

[0031] The preset rowing training time series data is a set of n triple data items in the form of sequence, type and value, arranged in the order of discrete time. The corresponding expression is:

[0032] T={(t1, T1, q1), (t2, T2, q2), ..., (t n , T j ,q n )} (8)

[0033] Where n represents the length of the sequence, t i (1≤i≤n) represents the time of the sequence, T j(1≤j≤28) indicates the data type of the sequence, q i (1≤i≤n) represents the data value of the sequence, (t i , T j ,q i )(1≤i≤n, 1≤j≤28) represents an element in a sequence;

[0034] The finite sequence set consisting of M rowing training time series data is:

[0035]

[0036] Among them, the length of the time series n x Indicates the length of the training time series data of the xth athlete, represents the yth training item of the xth athlete, Indicates the corresponding Measurement data for the project;

[0037] According to the rowing training time series data set, the minimum support min_sup and the minimum confidence min_con are set. If Sup(A→B)≥min_sup and Con(A→B)≥min_con, the association rule A→B is a strong association rule, in which the data items and time contained in A and B meet the constraints of the ternary data items.

[0038] As a preferred embodiment of the above technical solution, 26 training items are selected from 28 training items, and the remaining two training items are ignored. Taking an athlete x as an example, the initial value of each training item of the athlete is first recorded. One scan data record;

[0039] When you encounter the recorded data of a training project again, Calculate the difference with the corresponding initial value to get the result value d k (T j ), the corresponding expression is:

[0040]

[0041] The current project data Set to initial value, and continue scanning data records until the end;

[0042] After calculating the time series data difference d of each training item k (T j ), the value is judged to be rising, falling or flat according to the range of the difference, and is represented by corresponding symbols, that is, the changes in the athlete's training data are represented as a series of letter sequences represented by symbols, and the letter sequences are converted into a thing data set according to the sliding time window and the time window is set.

[0043] As a preferred embodiment of the above technical solution, a regression analysis algorithm is used based on rowing association rules and dynamics simulation model to predict rowing results to complete rowing simulation training, including:

[0044] The performance prediction is performed through the multivariate linear regression model algorithm. The overall model of the linear regression equation can be expressed by expression (10):

[0045] S=β0+β1X1+...+β m X m +e (10)

[0046] Among them, X represents the predicted water performance, X i (i=1, 2...m) represents the data of each rowing training project, β i (i=1, 2...m) represents the parameters of the model, and e represents the random error.

[0047] As a preferred embodiment of the above technical solution, light emitted by a light source on the water surface where the unmanned boat is located, reflected light from the surface of the water object, and refracted light from the bottom of the water are obtained, and an optical model of the water surface environment is constructed, including:

[0048] The expression of the optical model is: C = C rl ×R+C rt ×K, where C represents the comprehensive illumination of the water surface, C rl Indicates the contribution of the emitted light in the scene, C rt Represents the color contributed by the refracted light, R and K represent the reflection coefficient and refraction coefficient respectively and satisfy R+K=1. The calculation expression is:

[0049]

[0050] Among them, θ r represents the reflection angle, θ t represents the refraction angle;

[0051] During the lighting calculation process, the contribution value C of the refracted light rt As a constant, the contribution of reflected light C rl Consider the reflected light contribution from the ambient light and the reflected light contribution from the light source, that is, C rl =R(C sp +C ev ), where C sp Represents the reflected light from the light source, C ev Calculated value of sky reflected light;

[0052] The illumination value of each mesh face in the scene can be calculated by combining the normal vector and the incident angle of the light. Each face is rendered using a shader to obtain a realistic simulation scene.

[0053] In a second aspect, the present invention further provides a rowing simulation training system, comprising:

[0054] A first data acquisition unit is used to acquire light emitted by a light source on the water surface where the unmanned boat is located, reflected light from the surface of the water object, and refracted light from the bottom of the water, and to construct an optical model of the water surface environment;

[0055] The second data acquisition unit is used to acquire the position and posture information sent back by the unmanned boat when sailing on the water surface, and to construct a motion model of the unmanned boat according to the position and posture information and the optical model;

[0056] A model building unit, used for performing dynamic simulation on the racing boat according to the unmanned boat motion model to determine the dynamic simulation model of the racing boat;

[0057] A data processing unit, used to obtain historical rowing training data and pre-process the training data to obtain rowing association rules, wherein the historical training data includes data records of rowing athletes' daily training and testing, and rowing training time series data;

[0058] The simulation training unit is used to predict the rowing results based on the rowing association rules and dynamic simulation model using a regression analysis algorithm to complete the rowing simulation training.

[0059] In a third aspect, the present invention further provides a storage medium storing a computer program, which implements the steps of the above-mentioned rowing simulation training method when executed by a processor.

[0060] The present invention provides a rowing simulation training method, system and storage medium. The method comprises the following steps: obtaining light emitted by a light source on the water surface where an unmanned boat is located, reflected light from the surface of a water object and refracted light from the bottom of the water, and constructing an optical model of the water surface environment; obtaining position information sent back by the unmanned boat when sailing on the water surface; constructing an unmanned boat motion model according to the position information and the optical model; performing dynamic simulation on the rowing according to the unmanned boat motion model to determine the dynamic simulation model of the rowing; obtaining rowing training data and preprocessing the training data to obtain rowing association rules; using a regression analysis algorithm to predict rowing results based on the rowing association rules and the dynamic simulation model to complete rowing simulation training; setting the unmanned boat to collect light according to environmental factors and the rowing's own factors; establishing an optical model of the training environment; and combining the optical model with the unmanned boat motion model to construct the actual rowing motion process. The method can effectively provide a more realistic simulation vision of the rowing, improve the accuracy of real-time collection of training data, and automatically generate a training plan and training result prediction by analyzing the rowing association rules, effectively avoiding problems such as overtraining or undertraining of athletes, and improving the training level. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0062] Figure 1 A flow chart of the rowing simulation training method provided by the present invention;

[0063] Figure 2 This is a structural block diagram of the rowing simulation training system provided by the present invention. DETAILED DESCRIPTION

[0064] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0065] See also Figure 1 The present invention provides a rowing simulation training method, comprising the following steps:

[0066] S1: Obtain the light emitted by the light source on the water surface where the unmanned boat is located, the reflected light from the surface of the water object and the refracted light from the bottom of the water, and construct an optical model of the water surface environment;

[0067] S2: Obtain the posture information sent back by the unmanned boat when sailing on the water surface, and build the motion model of the unmanned boat based on the posture information and the optical model;

[0068] S3: Performing dynamic simulation on the racing boat according to the unmanned boat motion model to determine the dynamic simulation model of the racing boat;

[0069] S4: Acquire historical rowing training data and pre-process the training data to obtain rowing association rules, wherein the historical training data includes data records of rowing athletes' daily training and testing, and rowing training time series data;

[0070] S5: Based on the rowing association rules and dynamic simulation model, a regression analysis algorithm is used to predict rowing results to complete rowing simulation training.

[0071] In this embodiment, the position and posture information transmitted back by the unmanned boat when sailing on the water surface is obtained, and the motion model of the unmanned boat is constructed according to the position and posture information and the optical model, including: initializing various parameters of the unmanned boat, performing cyclic calculation according to the input propeller angle and engine speed, wherein the cyclic calculation includes calculating the fluid dynamics and torque, propeller thrust and torque, and the force and torque of wind and wave interference; constructing the inertial coordinate system and the attached coordinate system when the unmanned boat moves on the water surface, and dividing the displacement and rotation of the unmanned boat into three degrees of freedom; according to the corresponding relationship between the physical quantities of the inertial coordinate system and the attached coordinate system, two sets of expressions (1) and (2) are obtained;

[0072]

[0073] Among them, the physical quantities belonging to the inertial coordinate system are: the forces X0, Y0 and moment N0 acting on the boat, the coordinates x of the center of gravity G of the unmanned boat G0 ,y G0 ; The attached coordinate system includes: the forces X, Y on the hull, and the external torque. The x, Y, V, and r represent the forward speed, drift speed, and bow angular velocity of the unmanned boat, respectively. c Indicates the x-axis coordinate value of the center of the unmanned boat;

[0074] It should be noted that the unmanned boat in the preset inertial coordinate system is a rigid body. During the force movement, its shape, size and relative positions of different points in the boat body do not change. According to Newton's second law:

[0075]

[0076] Among them, m and I zz They represent the mass of the unmanned boat and the mass moment of inertia of the unmanned boat around the x-axis, I zz Expressed as L, B, C bRespectively represent the length, width, average draft and square coefficient of the unmanned boat, x' G0 ,y' G0 They represent the linear acceleration of the unmanned boat along the x0 and y0 directions respectively, and ψ' represents the bow acceleration of the unmanned boat;

[0077] After taking derivatives of the variables in expression (2), the variables are introduced into the basic motion equation of expression (3) and projected into the attached coordinate system. After calculation, the relational expression of the unmanned boat motion model is derived as follows:

[0078]

[0079] Among them, u', v', and r' respectively represent the forward speed, drift speed, and bow angular velocity in the unmanned boat motion model. Expanding expression (4) to obtain the complete unmanned boat motion model relationship expression is:

[0080]

[0081] Among them, the subscript H0 represents the fluid dynamic force and torque, the subscript P represents the propeller thrust and torque, the subscript WIND and WAVE represent the force and torque generated by wind and waves, and the correction term Y·x c Credit middle.

[0082] It should be understood that when conducting mathematical modeling of unmanned boats, the fluid dynamics and torques used need to be dimensionless to facilitate the comparison of corresponding analyses. According to fluid mechanics, the inertial fluid dynamics on the hull can be understood as follows: the unmanned boat is sailing in an ideal fluid, and the velocity of the fluid close to the hull will be affected by the motion state of the hull, so that the fluid produces a reverse force on the unmanned boat, which is the inertial fluid dynamics. The accuracy of the calculation of viscous fluid dynamics will affect the accuracy of the entire dynamic model. The general distributed method tries to perform regression analysis on the data collected multiple times, so as to obtain an approximate expression describing the relationship between the fluid dynamics derivative and the overall parameters, so as to estimate it. By acquiring the light emitted by the light source on the water surface where the unmanned boat is located, the reflected light from the surface of the water object and the refracted light from the bottom of the water, and constructing an optical model of the water surface environment, the position information sent back by the unmanned boat when sailing on the water surface is acquired, and the unmanned boat motion model is constructed according to the position information and the optical model, the rowing is dynamically simulated according to the unmanned boat motion model to determine the rowing dynamics simulation model, the rowing training data is acquired and the training data is preprocessed to obtain the rowing association rules, and the rowing results are predicted based on the rowing association rules and the dynamics simulation model using a regression analysis algorithm to complete the rowing simulation training, the unmanned boat is set to collect light according to environmental factors and the rowing's own factors, an optical model of the training environment is established, and the actual rowing motion process is constructed by combining the optical model and the unmanned boat motion model, which can effectively provide a more realistic simulation vision of the rowing and improve the accuracy of real-time collection of training data. The analysis of the rowing association rules can automatically generate a training plan, effectively avoid problems such as excessive or insufficient training of athletes, and improve the training level.

[0083] Optionally, the fluid power exerted on the unmanned boat is the reaction force generated by the fluid moving in the fluid on the surface of the boat body in contact with the unmanned boat, and the reaction force includes inertial force and viscous force. This breaks down into:

[0084]

[0085] Among them, the subscript l represents the inertial force, and H represents the viscous force;

[0086] The thrust vector model is used to divide the force generated by the propeller into thrust T and lateral pulling force N on the unmanned boat during steering. The preset propeller turning angle is δ, and the longitudinal and lateral thrusts of the propeller on the unmanned boat are:

[0087]

[0088] Among them, t P represents the thrust derating factor, α P represents the lateral force coefficient, X PIndicates the Y-axis offset of the center of the thruster in the attached coordinate system.

[0089] In this embodiment, rowing training data is obtained and preprocessed to obtain rowing association rules, including:

[0090] The preset rowing training time series data is a set of n triple data items in the form of sequence, type and value, arranged in the order of discrete time. The corresponding expression is:

[0091] T={(t1, T1, q1), (t2, T2, q2), ..., (t n , T j ,q n )} (8)

[0092] Where n represents the length of the sequence, t i (1≤i≤n) represents the time of the sequence, T j (1≤j≤28) indicates the data type of the sequence, q i (1≤i≤n) represents the data value of the sequence, (t i , T j ,q i )(1≤i≤n, 1≤j≤28) represents an element in a sequence;

[0093] The finite sequence set consisting of M rowing training time series data is:

[0094]

[0095] Among them, the length of the time series n x Indicates the length of the training time series data of the xth athlete, represents the yth training item of the xth athlete, Indicates the corresponding Measurement data for the project;

[0096] According to the rowing training time series data set, the minimum support min_sup and the minimum confidence min_con are set. If Sup(A→B)≥min_sup and Con(A→B)≥min_con, the association rule A→B is a strong association rule, in which the data items and time contained in A and B meet the constraints of the ternary data items.

[0097] It should be noted that 26 training items are selected from 28 training items, and the remaining two training items are ignored. Taking an athlete x as an example, the initial value of each training item of the athlete is first recorded. Scan the data once; when you encounter the recorded data of a training item again, Calculate the difference with the corresponding initial value to get the result value dk (T j ), the corresponding expression is:

[0098]

[0099] The current project data Set to initial value, and continue scanning data records until the end;

[0100] After calculating the time series data difference d of each training item k (T j ), the value is judged to be rising, falling or flat according to the range of the difference, and is represented by corresponding symbols, that is, the changes in the athlete's training data are represented as a series of letter sequences represented by symbols, and the letter sequences are converted into a thing data set according to the sliding time window and the time window is set.

[0101] It should be understood that after obtaining the required association rules, the effectiveness evaluation of these rules becomes an important indicator for applying these rules. In the initial evaluation stage, the support and confidence of the association rules are used as the criteria for judging the interest of the rules. After converting the rowing training data and generating the association rules, the rowing training association rules are obtained, which effectively improves the efficiency of the rowing simulation training.

[0102] Optionally, a rowing performance prediction is performed using a regression analysis algorithm based on rowing association rules and a dynamics simulation model to complete rowing simulation training, including:

[0103] The performance prediction is performed through the multivariate linear regression model algorithm. The overall model of the linear regression equation can be expressed by expression (10):

[0104] S=β0+β1X1+…+β m X m +e (10)

[0105] Among them, X represents the predicted water performance, X i (i=1, 2...m) represents the data of each rowing training project, β i (i=1, 2...m) represents the parameters of the model, and e represents the random error.

[0106] In this embodiment, light emitted by a light source on the water surface where the unmanned boat is located, reflected light from the surface of the water object, and refracted light from the bottom of the water are obtained, and an optical model of the water surface environment is constructed, including:

[0107] The expression of the optical model is: C = C rl ×R+C rt ×K, where C represents the comprehensive illumination of the water surface, C rl Indicates the contribution of the emitted light in the scene, C rtRepresents the color contributed by the refracted light, R and K represent the reflection coefficient and refraction coefficient respectively and satisfy R+K=1. The calculation expression is:

[0108]

[0109] Among them, θ r represents the reflection angle, θ t represents the refraction angle;

[0110] During the lighting calculation process, the contribution value C of the refracted light rt As a constant, the contribution of reflected light C rl Consider the reflected light contribution from the ambient light and the reflected light contribution from the light source, that is, C rl =R(C sp +C ev ), where C sp Represents the reflected light from the light source, C ev Calculated value of sky reflected light;

[0111] The illumination value of each mesh face in the scene can be calculated by combining the normal vector and the incident angle of the light. Each face is rendered using a shader to obtain a realistic simulation scene, thereby providing a more realistic visual experience for rowing simulation training.

[0112] See also Figure 2 The present invention also provides a rowing simulation training system, comprising:

[0113] A first data acquisition unit is used to acquire light emitted by a light source on the water surface where the unmanned boat is located, reflected light from the surface of the water object, and refracted light from the bottom of the water, and to construct an optical model of the water surface environment;

[0114] The second data acquisition unit is used to acquire the position and posture information sent back by the unmanned boat when sailing on the water surface, and to construct a motion model of the unmanned boat according to the position and posture information and the optical model;

[0115] A model building unit, used for performing dynamic simulation on the racing boat according to the unmanned boat motion model to determine the dynamic simulation model of the racing boat;

[0116] A data processing unit, used to obtain historical rowing training data and pre-process the training data to obtain rowing association rules, wherein the historical training data includes data records of rowing athletes' daily training and testing, and rowing training time series data;

[0117] The simulation training unit is used to predict the rowing results based on the rowing association rules and dynamic simulation model using a regression analysis algorithm to complete the rowing simulation training.

[0118] In this embodiment, in the multivariate regression model, not all independent variables have a significant relationship with the dependent variable. Therefore, it is necessary to screen from the independent variable set, introduce the independent variables that have a significant impact on the dependent variable into the equation, and remove the independent variables that have an insignificant impact from the equation. Considering the computational complexity of the problem application, the number of independent variables may be small, and stepwise regression is an effective method to achieve this goal. The basic idea of ​​stepwise regression is: first select an initial set containing some independent variables, then introduce a variable that has the greatest impact on the dependent variable from the variables outside the set in turn, then test the variables in the set, remove a variable with the smallest impact from the variables with insignificant impact, and repeat until the variable cannot be introduced and derived, and the introduction and derivation are based on a given significance level. When this process is completed, all independent variables in the regression equation are significant to the dependent variable, and the variables not in the equation are insignificant, which improves the accuracy of real-time collection of training data. Analysis of rowing association rules can automatically generate training plans and training performance predictions, effectively avoid problems such as overtraining or undertraining of athletes, and improve training levels.

[0119] In a feasible embodiment, the present invention further provides a storage medium storing a computer program, which implements the steps of the above-mentioned rowing simulation training method when executed by a processor.

[0120] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limiting, and thus other examples of the exemplary embodiments may have different values.

[0121] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0122] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A rowing simulation training method, characterized in that: The following steps are involved: Obtain the light emitted by the light source on the water surface where the unmanned boat is located, the reflected light from the surface of the water object, and the refracted light from the bottom of the water, and build an optical model of the water surface environment; Obtain the posture information sent back by the unmanned boat while sailing on the water surface, and build the unmanned boat motion model based on the posture information and the optical model; Perform dynamic simulation on the racing boat according to the unmanned boat motion model to determine the dynamic simulation model of the racing boat; Acquire historical rowing training data and pre-process the training data to obtain rowing association rules, wherein the historical training data includes data records of rowing athletes' daily training and testing, and rowing training time series data; Based on the rowing association rules and dynamic simulation model, a regression analysis algorithm is used to predict rowing results to complete rowing simulation training.

2. The rowing simulation training method according to claim 1, characterized in that: Obtain the posture information sent back by the unmanned boat when sailing on the water surface, and build the unmanned boat motion model based on the posture information and optical model, including: Initialize the parameters of the unmanned boat and perform cyclic calculations based on the input propeller angle and engine speed. The cyclic calculations include calculating the fluid dynamics and torque, propeller thrust and torque, and wind and wave interference forces and torques. Construct the inertial coordinate system and attached coordinate system of the unmanned boat when it moves on the water surface, and divide the displacement and rotation of the unmanned boat into three degrees of freedom; According to the corresponding relationship between the physical quantities of the inertial coordinate system and the attached coordinate system, two sets of expressions (1) and (2) are obtained; Among them, the physical quantities belonging to the inertial coordinate system are: the forces X0, Y0 and moment N0 acting on the boat, the coordinates x of the center of gravity G of the unmanned boat G0 ,y G0 ; The attached coordinate system includes: the forces X, Y on the hull, and the external torque. The x, Y, V, and r represent the forward speed, drift speed, and bow angular velocity of the unmanned boat, respectively. c Indicates the x-axis coordinate value of the center of the unmanned boat.

3. The rowing simulation training method according to claim 1, characterized in that: Also includes: The unmanned boat in the preset inertial coordinate system is a rigid body. During the force movement, its shape, size and relative positions of different points in the boat body do not change. According to Newton's second law: Among them, m and I zz They represent the mass of the unmanned boat and the mass moment of inertia of the unmanned boat around the x-axis, I zz Expressed as L, B, C b Respectively represent the length, width, average draft and square coefficient of the unmanned boat, x' G0 ,y' G0 They represent the linear acceleration of the unmanned boat along the x0 and y0 directions respectively, and ψ' represents the bow acceleration of the unmanned boat; After taking derivatives of the variables in expression (2), the variables are introduced into the basic motion equation of expression (3) and projected into the attached coordinate system. After calculation, the relational expression of the unmanned boat motion model is derived as follows: Among them, u', v', and r' respectively represent the forward speed, drift speed, and bow angular velocity in the unmanned boat motion model. Expanding expression (4) to obtain the complete unmanned boat motion model relationship expression is: Among them, the subscript H0 represents the fluid dynamic force and torque, the subscript P represents the propeller thrust and torque, the subscript WIND and WAVE represent the force and torque generated by wind and waves, and the correction term Y·x c Credit middle.

4. The rowing simulation training method according to claim 3, characterized in that: The fluid dynamics on the unmanned boat are the reaction forces generated by the fluid moving in the fluid on the surface of the boat body in contact with it. The reaction forces include inertial force and viscous force. This breaks down into: Among them, the subscript l represents the inertial force, and H represents the viscous force; The thrust vector model is used to divide the force generated by the propeller into thrust T and lateral pulling force N on the unmanned boat during steering. The preset propeller turning angle is δ, and the longitudinal and lateral thrusts of the propeller on the unmanned boat are: Among them, t P Indicates thrust derating factor, a P represents the lateral force coefficient, X P Indicates the Y-axis offset of the center of the thruster in the attached coordinate system.

5. The rowing simulation training method according to claim 1, characterized in that: The rowing training data is obtained and preprocessed to obtain the rowing association rules, including: The preset rowing training time series data is a set of n triple data items in the form of sequence, type and value, arranged in the order of discrete time. The corresponding expression is: T={(t1,T1,q1),(t2,T2,q2),...,(t n ,T j ,q n )} (8) Where n represents the length of the sequence, t i (1≤i≤n) represents the time of the sequence, T j (1≤j≤28) indicates the data type of the sequence, q i (1≤i≤n) represents the data value of the sequence, (t i ,T j ,q i )(1≤i≤n,1≤j≤28) represents the elements in the sequence; The finite sequence set consisting of M rowing training time series data is: Among them, the length of the time series n x Indicates the length of the training time series data of the xth athlete, represents the yth training item of the xth athlete, Indicates the corresponding Measurement data for the project; According to the rowing training time series data set, the minimum support min_sup and the minimum confidence min_con are set. If Sup(A→B)≥min_sup and Con(A→B)≥min_con, the association rule A→B is a strong association rule, in which the data items and time contained in A and B meet the constraints of the ternary data items.

6. The rowing simulation training method according to claim 5, characterized in that: Also includes: Select 26 out of 28 training items and ignore the remaining two training items. Take an athlete x as an example and first record the initial value of each training item of the athlete. One scan data record; When you encounter the recorded data of a training project again, Calculate the difference with the corresponding initial value to get the result value d k (T j ), the corresponding expression is: The current project data Set to initial value, and continue scanning data records until the end; After calculating the time series data difference d of each training item k (T j ), the value is judged to be rising, falling or flat according to the range of the difference, and is represented by corresponding symbols, that is, the changes in the athlete's training data are represented as a series of letter sequences represented by symbols, and the letter sequences are converted into a thing data set according to the sliding time window and the time window is set.

7. The rowing simulation training method according to claim 1, characterized in that: Based on the rowing association rules and dynamics simulation model, the rowing performance prediction is carried out using regression analysis algorithm to complete the rowing simulation training, including: The performance prediction is performed through the multivariate linear regression model algorithm. The overall model of the linear regression equation can be expressed by expression (10): S=β0+β1X1+...+β m X m +e (10) Among them, X represents the predicted water performance, X i (i=1,2...m) represents the data of each rowing training project, β i (i=1,2...m) represents the parameters of the model, and e represents the random error.

8. The rowing simulation training method according to claim 1, characterized in that: Obtain the light emitted by the light source on the water surface where the unmanned boat is located, the reflected light from the surface of the water object, and the refracted light from the bottom of the water, and build an optical model of the water surface environment, including: The expression of the optical model is: C = C rl ×R+C rt ×K, where C represents the comprehensive illumination of the water surface, C rl Indicates the contribution of the emitted light in the scene, C rt Represents the color contributed by the refracted light, R and K represent the reflection coefficient and refraction coefficient respectively and satisfy R+K=1. The calculation expression is: Among them, θ r Reflection angle, θ t represents the refraction angle; During the lighting calculation process, the contribution value C of the refracted light rt As a constant, the contribution of reflected light C rl Consider the reflected light contribution from the ambient light and the reflected light contribution from the light source, that is, C rl =R(C sp +C ev ), where C sp Represents the reflected light from the light source, C ev Calculated value of sky reflected light; The illumination value of each mesh face in the scene can be calculated by combining the normal vector and the incident angle of the light. Each face is rendered using a shader to obtain a realistic simulation scene.

9. A rowing simulation training system according to the rowing simulation training method according to any one of claims 1 to 8, characterized in that: include: A first data acquisition unit is used to acquire light emitted by a light source on the water surface where the unmanned boat is located, reflected light from the surface of the water object, and refracted light from the bottom of the water, and to construct an optical model of the water surface environment; The second data acquisition unit is used to acquire the position and posture information sent back by the unmanned boat when sailing on the water surface, and to construct a motion model of the unmanned boat according to the position and posture information and the optical model; A model building unit, used for performing dynamic simulation on the racing boat according to the unmanned boat motion model to determine the dynamic simulation model of the racing boat; A data processing unit, used to obtain historical rowing training data and pre-process the training data to obtain rowing association rules, wherein the historical training data includes data records of rowing athletes' daily training and testing, and rowing training time series data; The simulation training unit is used to predict the rowing results based on the rowing association rules and dynamic simulation model using a regression analysis algorithm to complete the rowing simulation training.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the rowing simulation training method according to any one of claims 1 to 8 are implemented.

Images