Automatic operation device for construction machine

Abstract

The automatic operation device includes a position estimation section having an interaction model that calculates actual position data corresponding to input force data as estimated position data; a force calculation section having a force calculation model that calculates force data corresponding to input deviation, and inputs the force data to the position estimation section; a database that stores past calculated parameters, i.e., basic parameters; and a parameter setting section that calculates object parameters corresponding to norms y(t) of actual positions and norms u(t) of forces from the basic parameters, and sets the object parameters as parameters of the interaction model and the force calculation model.

Description

Technical Field

[0001] The present invention relates to a technology for automatically operating engineering machinery. Background Art

[0002] In recent years, automatic operation devices have become known that allow hydraulic excavators to automatically perform a series of operations from excavation to dumping. For example, Patent Document 1 discloses a technology in which the engine speed of an automatic excavator is set for each arbitrary operation within a cycle of operations, by sequentially reading a taught and stored teaching position and repeatedly performing operations from excavation to dumping.

[0003] In the automatic operation of a hydraulic excavator, it is required to move the distal end of the attachment along a predetermined target trajectory regardless of the work site.

[0004] However, the properties of the soil excavated by the hydraulic excavator vary depending on the work site. Therefore, when moving the distal end of the attachment along a target trajectory, the hydraulic excavator must generate an appropriate excavation force that takes the properties of the soil into consideration.

[0005] The technology of Patent Document 1 does not take the characteristics of soil into consideration at all, and therefore cannot cause the hydraulic excavator to generate an appropriate digging force according to the characteristics of soil.

[0006] Such issues also occur in construction machinery other than hydraulic excavators.

[0007] Prior art literature

[0008] Patent Literature

[0009] Patent Document 1: Japanese Patent Application Laid-Open No. 2001-32330 Summary of the Invention

[0010] The present invention is made to solve the above-mentioned problems, and its purpose is to provide an automatic operation device as follows: considering the characteristics of the interaction between the working device and the object, the construction machine generates an appropriate force to make the position of the part where the working device and the object interact consistent with the target position.

[0011] An automatic operation device according to one aspect of the present invention is an automatic operation device for an engineering machine having a working device including a portion that interacts with an object, and includes: an acquisition unit for acquiring actual position data representing the actual position of the portion; an estimation unit for estimating the estimated actual position data by inputting estimated force data into a first model, the first model being a model that defines the relationship between force data and the actual position data using a first parameter representing a characteristic of the interaction, the force data representing a force generated at the portion; and a calculation unit for calculating a difference between the estimated actual position data and the actual position data. a deviation of the target position data representing the target position of the part; a calculation unit that calculates the estimated force data by inputting the deviation into a second model, the second model being a model that uses the first parameter to define the relationship between the deviation and the force data used to make the actual position consistent with the target position; a setting unit that calculates a second parameter corresponding to the estimated actual position data and the estimated force data based on the first parameter calculated in the past, and sets the first parameter based on the second parameter; and an instruction value calculation unit that calculates an instruction value for the construction machinery based on the estimated force data.

[0012] According to this configuration, the construction machine can generate an appropriate force for aligning the position of the portion where the working implement and the object interact with each other with the target position, taking into account the characteristics of the interaction between the working implement and the object. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a block diagram showing an example of the configuration of an automatic operation device according to an embodiment of the present invention.

[0014] Figure 2 is a schematic diagram of the interaction model.

[0015] Figure 3 This is a graph showing changes in the norm of the actual position during excavation.

[0016] Figure 4 It is a schematic diagram of the force direction.

[0017] Figure 5 Yes Figure 1 A flowchart showing an example of the processing of the automatic operation device is shown.

[0018] Figure 6 This is a flowchart showing the details of the parameter setting process.

[0019] Figure 7 This is a flowchart showing an example of processing of a construction machine in response to a command value input from an automatic operation device.

[0020] Figure 8 It is a block diagram showing the structure of the automatic operation device according to the embodiment.

[0021] Figure 9 It is a diagram showing a control target in the embodiment.

[0022] Figure 10 This is a diagram showing the relationship between the coordinates of the distal end of the attachment and the coordinates of the target position in the embodiment.

[0023] Figure 11 This is a diagram showing an overview of the verification model.

[0024] Figure 12 This is a table showing the values ​​of various parameters used to construct the initial database.

[0025] Figure 13 is a graph showing the simulation results for a fixed parameter controller.

[0026] Figure 14 is a graph showing the simulation results for a fixed parameter controller.

[0027] Figure 15 It is a graph showing the simulation results of the embodiment.

[0028] Figure 16 It is a graph showing the simulation results of the embodiment. DETAILED DESCRIPTION

[0029] The following embodiments of the present invention are described with reference to the accompanying drawings. The following embodiments are merely specific examples of the present invention and do not limit the technical scope of the present invention.

[0030] Figure 1 This is a block diagram illustrating an example of the configuration of an automatic operation device 1 according to an embodiment of the present invention. Automatic operation device 1 automatically operates a construction machine 200. Construction machine 200 is a construction machine such as a hydraulic excavator, a crane, or a demolition machine. In the following description, construction machine 200 is described as a hydraulic excavator. However, this is only an example; construction machine 200 may be any type of construction machine as long as it includes a working mechanism that interacts with an object.

[0031] Construction machine 200 comprises a lower traveling structure; an upper revolving structure rotatably mounted to the lower traveling structure; a boom mounted to the upper revolving structure so it can be raised and lowered; an arm mounted to the boom so it can swing; and a bucket mounted to the arm so it can swing. The boom, arm, and bucket constitute a working mechanism. Construction machine 200 also includes hydraulic cylinders for raising and lowering the boom, swinging the arm, and swinging the bucket.

[0032] The automatic operation device 1 may be incorporated into a controller of an existing construction machine 200 , or may be incorporated into a computer having a communication device capable of wirelessly communicating with the construction machine 200 .

[0033] The automatic operation device 1 includes: an acquisition unit 10; a position estimation unit 20 (an example of an estimation unit); a deviation calculation unit 30 (an example of a calculation unit); a force calculation unit 40 (an example of a calculation unit); an instruction value calculation unit 50; a database 60; a parameter setting unit 70 (an example of a setting unit); a force direction calculation unit 80; a target position acquisition unit 90; and a memory 100.

[0034] The acquisition unit 10 acquires the actual position coordinates Xt(t) of the bucket's distal end from the construction machine 200. The construction machine 200 has a function for detecting the bucket's distal end coordinates based on the rotation angle of the upper slewing structure, the angle of the boom relative to the upper slewing structure, the angle of the arm relative to the boom, and the angle of the bucket relative to the arm. Therefore, the acquisition unit 10 can acquire the bucket's distal end coordinates detected by this function from the construction machine 200 as the actual position coordinates Xt(t).

[0035] The actual position coordinates Xt(t) are, for example, coordinates on a two-dimensional plane perpendicular to the ground, with the bucket's distal end as the origin. Specifically, the actual position coordinates Xt(t) are expressed as Xt(t) = [xt(t), yt(t)]. Here, t is time, xt(t) is the x-axis component of the actual position in the two-dimensional coordinate system, and yt(t) is the y-axis component of the actual position in the two-dimensional coordinate system. The x-axis is, for example, set to the longitudinal direction of the work device, and the y-axis is set to be perpendicular to the ground.

[0036] The distal end of the bucket is an example of a portion where the working device interacts with the object. The origin of the coordinate system of the two-dimensional plane is set, for example, at the starting position of the interaction between the bucket and the object. The interaction between the bucket and the object refers to the contact between the bucket and the object and the mutual application of force to each other. The construction machine 200 detects whether the interaction has started based on, for example, the cylinder pressure value of the hydraulic cylinder, and inputs a notification indicating the start of the interaction to the acquisition unit 10. In addition, when the construction machine 200 detects the end of the interaction, it inputs a notification indicating the end to the acquisition unit 10. Thus, the acquisition unit 10 can determine whether the construction machine 200 is interacting with the object. The object is, for example, sand and soil contained in the ground excavated by the bucket.

[0037] The acquisition unit 10 calculates the norm of the actual position, |Xt(t)| = y(t), based on the acquired coordinates Xt(t) of the actual position. The coordinates Xt(t) of the actual position and the norm y of the actual position are stored in the memory 100(t). The coordinates Xt(t) of the actual position and the norm y(t) of the actual position are examples of actual position data.

[0038] The position estimation unit 20 includes an interaction model 21 (an example of a first model). The interaction model 21 defines the relationship between the norm u(t) of the force generated at the bucket's distal end when the work device and the object interact, and the norm y(t) of the actual position of the bucket's distal end, using parameters representing the characteristics of the interaction between the work device and the object. The force norm u(t) is an example of force data.

[0039] The position estimating unit 20 inputs the force norm u(t) calculated by the force calculation unit 40 into the interaction model 21 and calculates the norm of the actual position y(t) corresponding to the force norm u(t) as the estimated position norm y^(t). The position estimating unit 20 stores the calculated estimated position norm y^(t) in the memory 100. The estimated position norm y^(t) is an example of estimated actual position data. The interaction model 21 is represented by Equation (6) described below.

[0040] As shown in formula (6), the interaction model 21 is a function of the norm y^(t) of the estimated position and the norm u(t) of the force. "A^" and "B^" on the left are represented by formulas (7) and (8) described later. Formula (7) contains coefficients represented by a^1(t), a^2(t),... Formula (8) contains coefficients represented by b^0(t), b^1(t),... These coefficients are parameters of the interaction model 21 (an example of the first parameter). In this embodiment, as described later, the control object is modeled by formula (26), so the parameters of the interaction model 21 include a^1(t), a^2(t), and b^0(t).

[0041] The parameters a^1(t), a^2(t), and b^0(t) are expressed by equations (27) to (29) described below. As shown in equations (27) to (29), a^1(t), a^2(t), and b^0(t) contain m(t), c(t), and k(t). m(t) is the mass of the interaction between the working device and the object, k(t) is the spring constant of the spring element, and c(t) is the viscosity coefficient of the damping element. These parameters directly represent the characteristics of the interaction between the working device and the object.

[0042] Therefore, the parameters a^1(t), a^2(t), and b^0(t) indirectly represent the characteristics of the interaction between the working device and the object, and the interaction model 21 reflects the characteristics of the interaction.

[0043] The deviation calculation unit 30 retrieves the norm y(t-1) of the actual position and the norm y^(t-1) of the estimated position from the memory 100 and calculates the difference by subtracting y^(t-1) from y(t-1). Furthermore, the deviation calculation unit 30 calculates the deviation e(t) by subtracting the difference calculated based on the norm |R(t)| (=r(t)) of the target position input from the target position acquisition unit 90 and inputs this deviation to the force calculation unit 40. The deviation calculation unit 30 retrieves y(t-1) and y^(t-1) from the memory 100 because y(t) and y^(t) have not yet been calculated at the stage of calculating the deviation e(t). t-1 represents the sample point immediately preceding t.

[0044] The force calculation unit 40 includes a force calculation model 41. The force calculation model 41 is a model that defines the relationship between the deviation e(t) and the norm u(t) of the force generated at the distal end of the bucket to match the actual position with the target position, using the same parameters as the interaction model 21.

[0045] The force calculation model 41 is expressed by the equation (3) described later.

[0046] As shown in equation (3), the force calculation model 41 is a function of the force norm u(t) and the deviation e(t). Furthermore, "Q^" on the right is represented by equation (4) described later. As shown in equation (4), "Q^" includes "A^" and "B^." "A^" and "B^" are represented by a^1(t), a^2(t), and b^0(t) as described above. Therefore, it can be seen that the force calculation model 41 is defined by the same parameters as the interaction model 21.

[0047] The force calculation unit 40 inputs the deviation e(t) calculated by the deviation calculation unit 30 into the force calculation model 41 and calculates the force norm u(t) corresponding to the deviation e(t). The force calculation unit 40 inputs the calculated force norm u(t) into the command value calculation unit 50, the position estimation unit 20, and the memory 100. The calculated force norm u(t) is an example of estimated force data.

[0048] The command value calculation unit 50 calculates the force vector Fr(t) based on the force norm u(t) calculated by the force calculation unit 40 and the force direction θf(t) calculated by the force direction calculation unit 80. The command value calculation unit 50 then inputs the force vector Fr(t) as a command value to the construction machine 200. The command value calculation unit 50 may calculate the force vector Fr(t) using equation (31) described below.

[0049] The database 60 stores one or more basic parameters θ(t) that were previously calculated by the parameter setting unit 70. Each basic parameter θ(t) includes [a^1(t), a^2(t), b^o(t)].

[0050] The parameter setting unit 70 calculates the required point φ based on the basic parameters θ(t) stored in the database 60. - (t) corresponds to the object parameter θnewc(t) (an example of the second parameter). The required point φ(t) is φ(t) = [y(t), y(t-1), y(t-2), u(t-1)]. That is, the required point φ(t) includes the norms y(t), y(t-1), y(t-2) of the actual position and the norm u(t-1) of the force data. The required point φ(t) represents the dynamics of the current interaction of the construction machine 200 reflecting the current interaction between the working device and the object. In addition, the parameter setting unit 70 stores the average parameter θnew(t) described later obtained in the process of calculating the object parameter θnewc(t) as the basic parameter θ(t) in the database 60.

[0051] The force direction calculation unit 80 calculates the direction θf(t) of the force generated at the distal end of the bucket based on the target position coordinates R(t) input from the target position acquisition unit 90 and the actual position coordinates Xt(t-1) acquired from the memory 100. The actual position coordinates Xt(t-1) at time t-1 are acquired because the actual position coordinates Xt(t) have not yet been calculated at this stage. The force direction calculation unit 80 can calculate the force direction θf(t) using equation (30).

[0052] The target position acquisition unit 90 acquires the target position coordinates R(t) = [rX(t), ry(t)] and inputs them into the force direction calculation unit 80. The target position is the target position of the bucket's distal end. In this embodiment, after the interaction occurs, the automatic operation device 1 automatically operates the construction machine 200 so that the bucket's distal end moves along a predetermined target trajectory. Therefore, the target position is a position on this target trajectory. This target trajectory can be input by a manager, for example.

[0053] The target position acquisition unit 90 calculates the norm r(t) of the target position based on the coordinates R(t) of the target position and inputs the norm r(t) to the deviation calculation unit 30 .

[0054] Memory 100 is composed of RAM or flash memory, and stores the coordinates Xt(t) of the actual position, the norm y(t) of the actual position, and the norm y^(t) of the estimated position. Since point φ(t) is required to include the norms y(t), y(t-1), and y(t-2) of the actual position up to two samples ago and the norm u(t-1) of the force up to one sample ago, memory 100 only needs to store the norms y(t), y(t-1), and y(t-2) of the actual position up to at least two samples ago and the norm u(t-1) of the force up to at least one sample ago. Furthermore, since the norm y^(t-1) of the estimated position up to one sample ago is used when calculating the deviation e(t), memory 100 only needs to store the norm y^(t-1) of the estimated position up to at least one sample ago.

[0055] exist Figure 1 In the embodiment, each component other than the memory 100 constituting the automatic operation device 1 is constituted by, for example, a processor. The processor may be constituted by a central processing unit or a dedicated circuit such as an ASIC.

[0056] Figure 2 is a schematic diagram of the interaction model 21. Figure 2 As shown in the left column of , interaction model 21 is constructed based on the assumption that bucket 201 is moving within a two-dimensional plane 202. Two-dimensional plane 202 is a plane along the longitudinal direction of the work mechanism and perpendicular to ground surface 203. Two-dimensional plane 202 has an xt axis along the longitudinal direction of the work mechanism and a yt axis perpendicular to ground surface 203. Furthermore, the origin 204 of two-dimensional plane 202 is set at the location where bucket 201 and ground surface 203 begin to interact.

[0057] like Figure 2 As shown in the right column of , the interaction model 21 is a spring-mass-damping model including a mass unit 211, a damping unit 212, and a spring unit 213 of the interaction between the working device and the object. The mass unit 211 is represented by the mass m(t) of the interaction between the working device and the object. The damping unit 212 is represented by the viscosity coefficient c(t). The spring unit 213 is represented by the spring constant k(t). The damping unit 212 and the spring unit 213 are connected in parallel. The mass unit 211 is connected in series with the parallel unit formed by connecting the damping unit 212 and the spring unit 213 in parallel. The motion equation of this spring-mass-damping model is represented by equations (23) to (25) described later. Therefore, the interaction model 21 is composed of the model shown in equation (6) calculated based on equations (23) to (25).

[0058] like Figure 2As shown in the right column of , when the work mechanism is operating on a two-dimensional plane 202, the force F(t) generated at the distal end of bucket 201 and the coordinates Xt(t) of the actual position of the distal end of the bucket are each represented two-dimensionally. In interaction model 21, these are represented by the norm of F(t), |F(t)|, and the norm of the estimated position, |Xt(t)| (=y(t)). In other words, interaction model 21 is a dimensionality-compression model in which the input and output variables are dimensionally compressed. By constructing interaction model 21 as a dimensionality-compression model, interaction model 21 is simplified.

[0059] Figure 3 This is a graph showing the change in the norm y(t) of the actual position during excavation. Figure 3 In the example shown in FIG, the distal end of the bucket contacts the ground 203 at origin 204 and then moves along trajectory 205. The norm y(t) of the actual position is the distance between origin 204 and the actual position. Therefore, as the digging action progresses, the norm y(t) of the actual position increases.

[0060] Since the interaction model 21 is dimensionally compressed in this manner, when operating the construction machine 200, the force direction θf(t) can be instructed to the construction machine 200 in addition to the force norm u(t). Therefore, the force direction calculation unit 80 calculates the force direction θf(t).

[0061] Figure 4 This is a schematic diagram of the force direction θf(t). The force calculation unit 40 calculates the force norm u(t) so that the actual position matches the target position as described above. Therefore, if the coordinates of the actual position at time t-1 are taken as Xt(t-1), the force direction θf(t) at time t will be from the actual position coordinates Xt(t-1) toward the target position coordinates R(t). Therefore, the force direction calculation unit 80 calculates the force direction θf(t) using the actual position coordinates Xt(t-1) and the target position coordinates R(t).

[0062] Figure 5 Yes Figure 1 Flowchart showing an example of the process of the automatic operation device 1. In step S1, the acquisition unit 10 detects whether the interaction between the working device and the object has started. Here, the acquisition unit 10 may determine that the interaction has occurred when a notification notifying the start of the interaction is received from the construction machine 200.

[0063] If the start of interaction is detected ("YES" in step S1), the process proceeds to step S2, and if the start of interaction is not detected ("NO" in step S1), the process waits in step S1.

[0064] In step S2 , the target position acquisition unit 90 acquires the coordinates R(t) of the target position. For example, the target position acquisition unit 90 may sequentially acquire points on the target trajectory stored in the memory 100 as the coordinates R(t) of the target position.

[0065] In step S3, the target position acquisition unit 90 calculates the norm r(t) of the target position based on the coordinate R(t) of the target position. The norm r(t) of the target position is the distance from the origin to the target position when the interaction start position is the origin.

[0066] In step S4 , the deviation calculation unit 30 acquires the norm y(t−1) of the actual position and the norm ŷ(t−1) of the estimated position from the memory 100 .

[0067] In step S5 , the deviation calculation unit 30 calculates the deviation e(t) using the norm r(t) of the target position, the norm y(t−1) of the actual position, and the norm ŷ(t−1) of the estimated position as described above.

[0068] In step S6, the force calculation unit 40 inputs the deviation e(t) into the force calculation model 41 and calculates the force norm u(t). At this time, the force calculation unit 40 calculates u(t) using the initial parameter values ​​or the parameter θnew(t) determined in the previous step.

[0069] In step S7 , the position estimating unit 20 inputs the norm u(t) of the force into the interaction model 21 and calculates the norm ŷ(t) of the estimated position.

[0070] In step S8 , the force direction calculation unit 80 acquires the coordinate Xt(t−1) of the actual position from the memory 100 .

[0071] In step S9 , the force direction calculation unit 80 substitutes the coordinates R(t) of the target position and the coordinates Xt(t-1) of the actual position into equation (30) to calculate the direction of the force θf(t).

[0072] In step S10 , the command value calculation unit 50 substitutes the norm u(t) of the force and the direction θf(t) of the force into equation (31) to calculate the force vector Fr(t).

[0073] In step S11 , the command value calculation unit 50 inputs the force vector Fr(t) as a command value to the construction machine 200 .

[0074] In step S12 , the acquisition unit 10 acquires the coordinates Xt(t) of the actual position calculated by the construction machine 200 from the construction machine 200 as an input of the response command value.

[0075] In step S13 , the acquisition unit 10 calculates the norm y(t) of the actual position based on the coordinates Xt(t) of the actual position.

[0076] In step S14 , the acquisition unit 10 stores the coordinates Xt(t) and the norm y(t) of the actual position in the memory 100 .

[0077] In step S15, the parameter setting unit 70 executes a parameter setting process. Details of the parameter setting process will be described later.

[0078] In step S16, the acquisition unit 10 determines whether the interaction has ended. Here, the acquisition unit 10 may determine that the interaction has ended if it receives a notification from the construction machine 200 notifying the end of the interaction. The end of the interaction means that the distal end of the bucket and the object are no longer in contact. If the interaction is determined to have ended ("Yes" in step S16), the process ends. If the interaction is determined not to have ended ("No" in step S16), the process returns to step S2.

[0079] In this way, Figure 5 In the flowchart of , the processing by the automatic operation device 1 is executed sequentially during the occurrence of interaction.

[0080] Figure 6 1 is a flowchart showing the details of the parameter setting process. In step S101 , the parameter setting unit 70 acquires the required point φ(t) from the memory 100 .

[0081] In step S102 , the parameter setting unit 70 calculates the distance d between the required point φ(t) and the basic parameter θ(t) using equation (18) described later (step S102 ).

[0082] In step S103 , the parameter setting unit 70 extracts k basic parameters from the basic parameters θ(t) stored in the database 60 in ascending order of the distance d.

[0083] In step S104, the parameter setting unit 70 calculates the weight wj of each of the extracted k basic parameters using equation (19).

[0084] In step S105 , the parameter setting unit 70 calculates the weighted average value of the extracted k basic parameters, namely, the average parameter θnew(t), using equation (20).

[0085] In step S106 , the parameter setting unit 70 stores the average parameter θnew(t) in the database 60 as the basic parameter θ(t).

[0086] In step S107, the parameter setting unit 70 modifies the average parameter θnew(t) using equation (21) to calculate the target parameter θnewc(t). This modification is performed to prevent the control performance from deteriorating due to a sudden change in the average parameter θnew(t).

[0087] In step S108, the parameter setting unit 70 sets the target parameter θnewc(t) as a parameter of the interaction model 21 and a parameter of the force calculation model 41. Thus, appropriate parameters are set in the interaction model 21 and the force calculation model 41 according to the current interaction.

[0088] In step S109, the parameter setting unit 70 sets the basic parameter θ stored in the database 60. - The basic parameter θ(t) whose distance dj from the average parameter θnew(t) is less than the specified value β is extracted as redundant data, and the redundant data is deleted from the database 60. The distance dj is expressed by the equation (22) described later. After step S109 is completed, the processing enters Figure 5 Step S16.

[0089] Figure 7 This is a flowchart showing an example of processing of the construction machine 200 in response to a command value input from the automatic operation device 1. In step S301, the controller of the construction machine 200 obtains a command value from the automatic operation device 1. The command value is the force vector Fr(t) calculated by the command value calculation unit 50.

[0090] In step S302, the controller of the construction machine 200 detects the posture of the working device. Here, the controller of the construction machine 200 detects the angle of the boom, the angle of the arm, and the angle of the bucket detected by the angle sensor as the posture of the working device.

[0091] In step S303, the controller of the construction machine 200 calculates the torque generated by the boom, arm, and bucket based on the posture of the work device and the specification data of the work device. The specification data includes, for example, the mass and length of the boom, arm, and bucket.

[0092] In step S304 , the controller of the construction machine 200 calculates the forces generated by the hydraulic cylinders of the boom, the arm, and the bucket based on the torques generated by the boom, the arm, and the bucket.

[0093] In step S305 , the controller of the construction machine 200 calculates command values ​​for the control valves of the boom, the arm, and the bucket based on the forces generated by the boom, the arm, and the bucket.

[0094] In step S306 , the controller of the construction machine 200 detects the coordinates Xt(t) of the actual position of the distal end of the bucket. The detected coordinates Xt(t) are input to the automatic operation device 1 .

[0095] Thus, according to the automatic operation device 1 of this embodiment, the target parameter θnewc(t) is calculated based on the previously calculated basic parameter θ(t). This target parameter θnewc(t) is set as a parameter of the interaction model 21 and the force calculation model 41. This target parameter θnewc(t) corresponds to the force norm u(t-1) calculated by the force calculation model 41 and the actual position norms y(t), y(t-1), and y(t-2) acquired by the acquisition unit 10. Furthermore, the force calculation model 41, to which this target parameter θnewc(t) is set, calculates the force norm u(t) required to align the bucket's distal end with the target position. Based on the calculated force norm u(t), a command value is calculated and input to the working device. Here, the correlation between the actual position norm y(t) and the force norm u(t) includes the characteristics of the interaction. Therefore, the target parameters corresponding to the actual position norm y(t) and the force norm u(t) reflect the characteristics of the interaction. This allows parameters reflecting the interaction characteristics to be set in the interaction model 21 and the force calculation model 41. As a result, the construction machine can generate appropriate force to match the position of the interaction portion to the target position in consideration of the interaction characteristics.

[0096] Furthermore, the above-described embodiment can adopt the following modified examples.

[0097] (1) The output variable of the force calculation model 41 and the input variable of the interaction model 21 are not limited to the force norm u(t) but may also be a two-dimensional or three-dimensional vector representing the force. In this case, the force direction calculation unit 80 is unnecessary; the command value calculation unit 50 can simply input the two-dimensional or three-dimensional vector representing the force as the command value to the construction machine 200.

[0098] (2) The output variable of the interaction model 21 is not limited to the norm y^(t) of the estimated position, but may be the two-dimensional coordinates or three-dimensional coordinates of the estimated position.

[0099] (3) The interaction model 21 is constructed by assuming that the bucket 201 moves on a two-dimensional plane 202. However, it can also be constructed by assuming that the bucket 201 moves on a three-dimensional plane. In this case, the interaction model 21 can be constructed taking into account not only the working mechanism but also the rotational movement of the upper rotating body.

[0100] (4) The interaction model 21 is a spring-mass-damper model, but any model may be used as long as it represents the relationship between force data and estimated position data.

[0101] (5) The interaction model 21 includes the damping element 212 and the spring element 213 , but either element may be omitted.

[0102] (6) The database 60 may store the object parameter θnewc(t) instead of the average parameter θnew(t). Furthermore, the database 60 may store the mass m(t), spring constant k(t), and viscosity coefficient c(t) of the interaction between the working device and the object as parameters. In this case, the parameter setting unit 70 may convert the mass m(t), spring constant k(t), and viscosity coefficient c(t) into parameters a^1(t), a^2(t), and b^0(t) using equations (27) to (29) described later. Furthermore, the parameter setting unit 70 may calculate the object parameter θnewc(t) using the converted parameters a^1(t), a^2(t), and b^0(t).

[0103] (7) The parameter setting unit 70 may set the average parameter θnew(t) as a parameter of the interaction model 21 and the force calculation model 41 instead of the target parameter θnewc(t). In this case, the average parameter θnew(t) is an example of the target parameter.

[0104] (8) When the construction machine 200 is composed of a dismantling machine including a crusher instead of a bucket, the force data may be data indicating the gripping force of the crusher of the dismantling machine to grip the object.

[0105] (9) The distal end of the bucket is used as the interaction site, but a position other than the distal end of the bucket (for example, the center of gravity or the center of the bucket) may be used as the interaction site.

[0106] (10) Figure 1 The construction machine 200 shown may also be a digital twin of the construction machine reproduced in a computer space rather than an actual construction machine.

[0107] (Example)

[0108] Next, embodiments of the present invention will be described. Figure 8 This is a block diagram showing the structure of the automatic operation device involved in the embodiment. The automatic operation device includes an internal model control system based on a database-driven method. In this embodiment, a mathematical model of a hydraulic excavator is used as the construction machine 200. This mathematical model is represented by equation (32) described below.

[0109] The automatic operation device involved in the embodiment includes a norm calculation unit 810, a subtraction unit 811, an internal model 820, a subtraction unit 830, a controller 840, a force vector calculation unit 850, a database 860, a parameter setting unit 870, a force direction calculation unit 880, and a norm calculation unit 890.

[0110] exist Figure 8 In, given and Figure 1 The part with the same name Figure 1 The internal model 820 corresponds to the interaction model 21 , and the controller 840 corresponds to the force calculation model 41 .

[0111] The norm calculation unit 810 corresponds to Figure 1 The acquisition unit 10 calculates the norm of the coordinate Xt(t) of the actual position. The subtraction unit 811 and the subtraction unit 830 correspond to Figure 1 The deviation calculation unit 30 is configured. The subtraction unit 811 calculates the difference between the norm y(t) of the actual position and the norm y^(t) of the estimated position. The subtraction unit 830 subtracts this difference from the norm |R(t)| of the target position to calculate the deviation e(t). The norm calculation unit 890 calculates the norm |R(t)| of the target position based on the coordinates R(t) of the target position.

[0112] The control object of the embodiment can be regarded as a discrete-time nonlinear system represented by equation (1).

[0113] y(t)=h(φ(t-1)) (1)

[0114] y(t) represents the output of the discrete-time nonlinear system, h(·) represents the nonlinear function, and φ(t-1) represents the information vector. The information vector φ(t-1) is defined as follows.

[0115] φ(t-1):=[y(t-1),···,y(tn y ), u(t-1), ···, u(tn u -1)] (2)

[0116] u(t) represents the input, ny and nu represent the number of output (y(t)) and input (u(t)) respectively.

[0117] Figure 1 The internal model control system shown can be expressed by the following equation.

[0118]

[0119]

[0120]

[0121]

[0122] r(t) represents the control target value, y^(t) represents the norm of the estimated position output from internal model 820, λ represents the filter design parameter, and n represents the filter order. Furthermore, A^(z-1, t) and B^(z-1, t) contain polynomials describing the discrete-time nonlinear system shown below. A^(z-1, t) and B^(z-1, t) are assumed to be locally stable minimum-phase systems.

[0123]

[0124]

[0125] The control object represented by equation (1) can be partially described by the following equation.

[0126]

[0127] At this time, when using equation (26) which models the controlled object, equation (9) is described as follows.

[0128] y(t)=θ(t)φ(t-1) T (10)

[0129]

[0130] φ(t-1): =[y(t-1), y(t-2), u(t-1)] (12)

[0131] According to equation (10), the parameter θ(t) is described as follows: The parameter θ(t) is a parameter of the discrete-time nonlinear system.

[0132]

[0133]

[0134] Here, f(·) represents a linear function. In order to calculate the local parameter θ(t) at each moment, the point φ is required - (t) and the basic parameters θ stored in the database 860 - (j) is defined as follows.

[0135]

[0136]

[0137] θ - (j) will be described in detail later.

[0138] The adjustment process of the parameters of the controller 840 and the internal model 820 based on the database-driven method is as follows.

[0139] [Step #1] Initial database construction

[0140] The parameter setting unit 870 uses the sequential least square method using the input and output data of the control object to obtain the parameters of equation (26). The parameter setting unit 870 uses the obtained parameters as basic parameters θ - The parameter setting unit 870 stores the basic parameter θ(j) in the initial database Θ defined by the following formula: - (j).

[0141]

[0142] NO indicates the number of basic parameters.

[0143] [Step #2] Calculation of system parameters

[0144] The parameter setting unit 870 calculates the required point φ using the following formula: - (t) and each basic parameter θ - (j) The parameter setting unit 870 sets each basic parameter θ - (j) Sort by distance from smallest to largest.

[0145]

[0146] Here, N(t) is given the required point φ - The number of basic parameters stored in the database 860 at time (t). i represents the i-th unit of the required point and basic parameter. Equation (18) represents the basic parameter θ - (j) and the required point φ on the hyperplane obtained by equation (9) - (t) The parameter setting unit 870 is the distance between - (t), θ - The smaller of (j)) extracts k basic parameters and calculates the weight wj of each basic parameter by the following formula.

[0147]

[0148] Here, nw is a design parameter for making the difference in weight corresponding to the distance significant. Furthermore, the parameter setting unit 870 calculates k basic parameters θ by the local linear averaging method shown in the following equation: - (t) and serves as the basic parameter θ - (t) Stored in database 860.

[0149]

[0150] [Step #3] Input Decision Preprocessing

[0151] In order to prevent the deterioration of control performance due to a sudden change in the average parameter θnew(t) obtained in step #2, the parameter setting unit 870 modifies the average parameter θnew(t) using a first-order lag filter represented by the following equation.

[0152]

[0153] α represents a filter design parameter, determined by trial and error. Parameter setting unit 870 uses the average parameter θnew(t) modified by equation (21) as the target parameter θnewc(t). Furthermore, parameter setting unit 870 applies the target parameter θnewc(t) to controller 840 shown in equation (3) and internal model 820 shown in equation (6).

[0154] [Step #4] Deletion of redundant data

[0155] Considering the storage capacity and computational cost of the implementation target, it is ideal to delete redundant data in the database 860. The parameter setting unit 870 deletes basic parameters that meet the following conditions from the basic parameters.

[0156]

[0157] β represents a design parameter for selecting a basic parameter to be deleted, and is determined by a trial-and-error method.

[0158] When there are multiple basic parameters that satisfy the condition of formula (22), the parameter setting unit 870 deletes only the most recent basic parameter.

[0159] By performing the processes from [Step #2] to [Step #4] at each time, the object parameter θnewc(t) reflecting the current interaction is calculated online. The parameter setting unit 870 applies the successively calculated object parameter θnewc(t) to the controller 840 and the internal model 820.

[0160] Next, the interaction model of the hydraulic excavator will be described.

[0161] The interaction model is a model that uses the interaction between the distal end of the hydraulic excavator's attachments (including the working device of the bucket) and the environment (object) as the control object. The hydraulic excavator operates by combining the movement of the attachments and the rotation movement of the main body. In this embodiment, the interaction model is constructed based on the movement of the attachments only. The interaction between the attachments and the environment can be assumed to be the resistance generated locally by the mass unit, spring unit, and damping unit. The control object can be Figure 9The equation of motion for this model is shown below.

[0162]

[0163] y(t)=|X t (t)| (24)

[0164] u(t)=|F(t)| (25)

[0165] Xt(t)=[xt(t),yt(t)] T Indicates the position of the remote end of the attached device. F(t) = [fx(t), fy(t)] T represents the force vector at the distal end of the attachment. m(t) represents the mass of the interaction between the working device and the object. k(t) represents the spring constant. c(t) represents the viscosity coefficient.

[0166] The interaction characteristics between the distal end of a hydraulic excavator's attachments and the environment vary depending on operating and environmental conditions. In this embodiment, these changes are represented by changes in the model parameters: mass m(t), spring constant k(t), and viscosity coefficient c(t) of the interaction between the working device and the object. By discretizing Equation (23) using the difference method, the discrete-time nonlinear system of the control target is obtained as shown below.

[0167]

[0168] According to formula (23), the parameters a^1(t), a^2(t), and b^0(t) are represented by the parameters of the interaction model, namely m(t), k(t), and c(t), as shown in the following formula.

[0169]

[0170]

[0171]

[0172] Ts is the sampling time.

[0173] Next, the direction θf(t) of the force generated at the distal end of the accessory device will be described.

[0174] Equation (23) is a scalar value representing the norm of force u(t). In order to control the hydraulic excavator, the direction of the force θf(t) is required. The direction of the force θf(t) is determined by Figure 10 The relationship between the coordinates Xt(t)=[xt(t), yt(t)]T of the distal end of the accessory device shown and the coordinates R(t)=[rx(t), ry(t)]T of the target position is determined by the following formula.

[0175]

[0176] Furthermore, based on u(t) calculated by equation (3) and equation (30), the force vector Fr(t) is determined by the following equation: Thus, the control of the hydraulic excavator is realized.

[0177] F r (t)=u(t)[sinθ f (t), cosθ f (t)] T (31)

[0178] Next, a simulation performed to verify the embodiment will be described.

[0179] In this simulation, a validation model with object jobs as mining was used. Figure 11 This diagram outlines the verification model. In the verification model, the attached equipment is considered a rigid two-link manipulator to simplify the structure. The equation of motion for the verification model is shown below.

[0180]

[0181] Here, τ(t) = [τ1(t), τ2(t)]T represents the joint torque at time t. Fre(t) represents the excavation reaction force. M(t) represents the inertia matrix. q(t) = [q1(t), q2(t)]T represents the joint angle. s(q·(t), q(t)) represents the squared velocity term and the gravity term. J(t) represents the Jacobian matrix. The excavation reaction force Fre(t) is calculated using the Rankine's passive earth pressure Frp(t) using the following equation.

[0182]

[0183] γs(t) represents the soil's weight per unit volume. h(t) represents the retaining wall height. ψs(t) represents the soil's internal friction angle. γs(t) and ψs(t) are parameters that vary depending on the soil quality. The retaining wall height h(t) is calculated based on the geometric relationship between the amount of soil in the bucket and the bucket angle. Assuming that the excavation reaction force Fre(t) is generated at the distal end of the bucket in a direction perpendicular to the bucket opening, the excavation reaction force Fre(t) is expressed as follows.

[0184] F re (t) = F rp (t)[sinθ re (t), cosθ re (t)] T (34)

[0185]

[0186] Secondly, the use of Figure 11 The initial database for the validation model shown in Figure 1 is constructed. First, the distal end of the bucket moves along a specified target trajectory. Here, joint torques are generated based on PD control, and the distal end of the manipulator follows. Figure 12 This table shows the various parameter values ​​used to construct the initial database. The parameters are calculated using the sequential least squares method based on time series data of the excavation force norm u(t) and the position norm y(t) of the manipulator's distal end relative to the excavation start point for each condition. The calculated parameters are stored as the initial database.

[0187] Next, the verification results are described.

[0188] The comparison results of the comparative example and the embodiment with fixed parameters are described below. Figure 12 The values ​​of the soil parameters are set as follows so that the soil quality changes with the excavation depth.

[0189]

[0190] y2th1 and y2th2 represent the coordinates of the far end of the auxiliary equipment for changing the soil parameters. Figure 13 、 Figure 14 It is a graph showing the simulation results of the comparative example. Figure 15 、 Figure 16 The following are graphs showing the simulation results of the embodiment. In these graphs, the norm u(t) of the force input to the hydraulic excavator is normalized with the maximum value being 100%. Figure 14 、 Figure 16 In the equation, X2(t) represented by “O” and R2(t) represented by “*” represent Figure 11 The coordinates of the distal end of the attachment and the target coordinates in the manipulator's coordinate system.

[0191] like Figure 14 As shown in , since the comparative example cannot express the characteristics of the control object that changes successively, the tracking performance of the target trajectory is poor. Figure 13 As shown in , the norm u(t) of the input force vibrates. On the other hand, as Figure 16 As shown, in the embodiment, Figure 15As shown, the parameters are calculated successively as the posture of the attachment and the soil quality change. Moreover, compared with the fixed parameter controller, the vibration of the input force norm u(t) is suppressed. When being assembled, it is more ideal that the input force norm u(t) obtains a stable value, so it can be seen that the embodiment is more suitable for assembly than the comparative example. In this verification, it was confirmed that the embodiment has a 61% improvement in tracking the target trajectory compared to the comparative example. In summary, it was confirmed that for unknown and time-varying work objects, the method of the embodiment can adapt to the changes of the work object and can achieve excavation that can track the target trajectory.

[0192] (Summary of Implementation Methods)

[0193] An automatic operation device according to one aspect of the present invention is an automatic operation device for an engineering machine having a working device including a portion that interacts with an object, and includes: an acquisition unit for acquiring actual position data representing the actual position of the portion; an estimation unit for estimating the estimated actual position data by inputting estimated force data into a first model, the first model being a model that defines the relationship between force data and the actual position data using a first parameter representing a characteristic of the interaction, the force data representing a force generated at the portion; and a calculation unit for calculating a difference between the estimated actual position data and the actual position data. a deviation of the target position data representing the target position of the part; a calculation unit that calculates the estimated force data by inputting the deviation into a second model, the second model being a model that uses the first parameter to define the relationship between the deviation and the force data used to make the actual position consistent with the target position; a setting unit that calculates a second parameter corresponding to the estimated actual position data and the estimated force data based on the first parameter calculated in the past, and sets the first parameter based on the second parameter; and an instruction value calculation unit that calculates an instruction value for the construction machinery based on the estimated force data.

[0194] According to this technical solution, a second parameter is calculated based on a previously calculated first parameter. This second parameter is set as the first parameter for the first and second models. This second parameter corresponds to the estimated force data calculated using the second model and the actual position data acquired by the acquisition unit. Furthermore, using the second model to which this first parameter is set, estimated force data for aligning the interacting portion with the target position is calculated. Based on the calculated estimated force data, a command value for the construction machine is calculated and input into the working device. Here, the correlation between the actual position data and the force data includes the characteristics of the interaction. Therefore, the first parameter corresponding to the actual position data and the estimated force data reflects the characteristics of the interaction. Thus, the first parameter reflecting the characteristics of the interaction can be set in the first and second models. As a result, the construction machine can generate an appropriate force to align the position of the interacting portion with the target position, taking into account the characteristics of the interaction.

[0195] In the automatic operation device described above, preferably, the estimated force data and the estimated actual position data are norms.

[0196] According to this aspect, since the output variable of the second model and the input and output variables of the first model are expressed in one dimension, the second model and the first model can be constructed using a simple model.

[0197] In the above-mentioned automatic operation device, it is more ideal that the actual position data and the target position data include coordinate data, and the automatic operation device also includes: a direction calculation unit, which calculates the direction of the force generated at the part based on the coordinate data represented by the actual position data and the coordinate data represented by the target position data; wherein the instruction value calculation unit calculates the force vector generated at the part based on the direction of the force and the norm of the estimated force data, and calculates the instruction value including the force vector.

[0198] According to this technical solution, the direction of the force generated at the interacting portion is calculated based on the coordinate data of the actual position and the coordinate data of the target position. A force vector is then calculated based on the calculated force direction and the norm of the estimated force data calculated by the calculation unit. A command value including the calculated force vector is then input to the construction machine. This allows the construction machine to be instructed not only on the magnitude but also on the direction of the force, enabling it to achieve appropriate movement.

[0199] In the automatic operation device described above, preferably, the first parameter is defined by the mass of the interaction and at least one of a spring constant and a viscosity coefficient representing the interaction.

[0200] According to this aspect, since the first parameter is defined by the interacting mass and at least one of the spring constant and the viscosity coefficient representing the interacting mass, the first model and the second model can more accurately reflect the characteristics of the interacting mass.

[0201] In the above-mentioned automatic operation device, it is more ideal that the acquisition unit obtains a notification indicating whether the interaction has started from the construction machinery, and the estimation unit, the calculation unit, the operation unit, the setting unit, and the instruction value calculation unit successively perform processing when the interaction occurs.

[0202] According to this technical solution, as the interaction occurs, the parameters are updated sequentially, so that the first parameters suitable for the characteristics of the sequentially changing interaction can be set in the first model and the second model, and the construction machine can generate a force suitable for the characteristics of the interaction.

[0203] In the automatic operation device described above, preferably, the calculation unit calculates a difference between a norm of the actual position data and a norm of the estimated actual position data, and a difference between a norm of the target position data as the deviation.

[0204] According to this technical solution, since the difference between the norm of the actual position data and the norm of the estimated position data and the difference between the norm of the target position data are calculated as deviations and input into the operation unit, the input variable of the second model, namely the deviation, can be constructed in one dimension, thereby simplifying the construction of the second model.

[0205] In the above-mentioned automatic operation device, it is preferable that the said part is the distal end of the said working device.

[0206] According to this aspect, it is possible to generate an appropriate force at the distal end of the working device so as to match the position of the distal end of the working device with the target position, taking into account the characteristics of the interaction.

[0207] In the automatic operation device described above, preferably, the construction machine is a hydraulic excavator, the object is sand and soil, and the force is an excavation force.

[0208] According to this aspect, the hydraulic excavator can generate an appropriate excavation force that causes the position of the distal end of the working device to coincide with the target position, taking into account the characteristics of the soil.

[0209] The automatic operation device described above preferably further includes a database storing the first parameter calculated in the past.

[0210] According to this aspect, since the database storing the first parameter calculated in the past is provided, it is easy to obtain the first parameter calculated in the past.

Claims

1. An automatic operation device for a construction machine, wherein the construction machine is provided with a working device including a portion for interacting with an object, characterized in that include: an acquiring unit for acquiring actual position data indicating the actual position of the part; an estimating unit that estimates estimated actual position data by inputting estimated force data into a first model, the first model being a model that defines a relationship between force data and the actual position data using a first parameter representing a characteristic of the interaction, the force data representing a force generated at the site; a calculation unit that calculates a deviation between a difference between the estimated actual position data and the actual position data and target position data indicating a target position of the part; a calculation unit that calculates the estimated force data by inputting the deviation into a second model that defines a relationship between the deviation and the force data for aligning the actual position with the target position using the first parameter; a setting unit that calculates a second parameter corresponding to the estimated actual position data and the estimated force data based on the first parameter calculated in the past, and sets the first parameter based on the second parameter; as well as, A command value calculation unit calculates a command value for the construction machine based on the estimated force data.

2. The automatic operation device according to claim 1, characterized in that: The estimated force data and the estimated actual position data are norms.

3. The automatic operation device according to claim 2, characterized in that: The actual position data and the target position data include coordinate data, The automatic operation device further includes: a direction calculation unit, which calculates the direction of the force generated at the part based on the coordinate data represented by the actual position data and the coordinate data represented by the target position data; wherein, The command value calculation unit calculates a force vector generated at the site based on the direction of the force and a norm of the estimated force data, and calculates the command value including the force vector.

4. The automatic operation device according to claim 2 or 3, characterized in that: The first parameter is defined by the mass of the interaction and at least one of a spring constant and a viscosity coefficient representing the interaction.

5. The automatic operation device according to any one of claims 2 to 4, characterized in that: The acquiring unit acquires a notification indicating whether the interaction has started from the construction machine, The estimating unit, the calculating unit, the arithmetic unit, the setting unit, and the command value calculating unit successively execute processes while the interaction occurs.

6. The automatic operation device according to any one of claims 2 to 5, characterized in that: The calculation unit calculates a difference between a norm of the actual position data and a norm of the estimated actual position data, and a difference between a norm of the target position data as the deviation.

7. The automatic operation device according to any one of claims 1 to 6, characterized in that: The location is the distal end of the working device.

8. The automatic operation device according to any one of claims 1 to 7, characterized in that: The engineering machinery is a hydraulic excavator, The object is sand, The force is the digging force.

9. The automatic operation device according to any one of claims 1 to 8, characterized in that Also includes: The database stores the first parameter calculated in the past.

Images