Operation control method, device and equipment of website robot, medium and program

Abstract

The invention discloses an operation control method, device and equipment of a network robot, a medium and a program. The method comprises the following steps: acquiring a predefined cost function; extracting a course angle state value in the cost function in real time, and determining a first error rate and a first error change rate of the basic domain according to an expected value; projecting a second error rate and a second error change rate of the cloud model discourse domain by using the discourse domain input conversion coefficient; determining a first qualitative concept and a second qualitative concept in the combined input normal cloud curve; in the course angle coefficient adjustment rule table, determining a qualitative concept output value, and projecting a conversion coefficient as an output value of a basic discourse domain according to the discourse domain; and combining the output value with an original course angle weight coefficient in the error weight matrix to obtain a new error weight, and performing operation control by using the new error weight. According to the embodiment of the invention, by continuously updating the error weight, the overall control effect and motion performance are improved, and powerful support is provided for intelligent decision and control based on the cloud model.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, applicable to financial technology scenarios, and in particular to an operation control method, device, equipment, medium and program for a branch robot. Background Art

[0002] Mobile robots used in bank branches can sense the driving environment, make autonomous decisions, plan driving paths, and control themselves to drive in the desired motion state, autonomously track the desired path, reach the set destination, and complete the scheduled task. The existing robot control scheme relies on traditional model predictive control (MPC), that is, through system modeling, model linearization, model discretization, model prediction, optimization of control input, feedback correction and other processes to output the robot's control signal, so as to achieve timely regulation.

[0003] In traditional MPC control, the motion mathematical model constructed based on the robot's motion state often contains multiple loss terms in its cost function, and the weight of each loss term is set based on human experience. However, the robot's motion state is greatly affected by the surrounding environment and the weight of each loss term cannot be flexibly adjusted, especially the weight of the error, which will directly affect the control accuracy of the robot.

[0004] Therefore, the embodiment of the present invention focuses on the heading angle in the robot's motion state, and optimizes the output result of the loss function by updating the weight corresponding to the heading angle, thereby further adapting to the changing driving environment and driving tasks. Summary of the invention

[0005] Based on this, the present invention provides an operation control method, device, equipment, medium and program for a network robot to solve the problem of poor robot control accuracy due to the inability to adaptively adjust the parameter weights in the predefined cost function in the process of outputting robot control signals using traditional model predictive control.

[0006] In a first aspect, an embodiment of the present invention provides an operation control method of a network point robot, the method comprising:

[0007] In the process of using the pre-built joint controller to control the operation of the network intelligent robot, the MPC cost function predefined in the joint controller is obtained;

[0008] Extract the heading angle state value in the mathematical model of the mesh intelligent robot motion in real time in the predefined MPC cost function, and determine the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle;

[0009] Using the domain input conversion coefficients corresponding to the first error rate and the first error change rate, respectively, the first error rate and the first error change rate are projected into the second error rate and the second error change rate of the heading angle in the cloud model domain;

[0010] Determining respectively a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve;

[0011] Determining a target qualitative concept corresponding to the heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept;

[0012] Determining an output value of a qualitative concept corresponding to a target qualitative concept in a pre-generated combined output normal cloud curve, and projecting the output value into an output value of a basic domain according to a domain output conversion coefficient;

[0013] The output value is combined with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and the new error weight is used to continue to control the operation of the network point intelligent robot.

[0014] In a second aspect, an embodiment of the present invention provides an operation control device for a network point robot, the device comprising:

[0015] The cost function acquisition module is used to acquire the MPC cost function predefined in the joint controller during the process of using the pre-structured joint controller to control the operation of the network intelligent robot;

[0016] The error calculation module is used to extract the heading angle state value in the mathematical model of the mesh intelligent robot movement in real time in the predefined MPC cost function, and determine the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle;

[0017] a domain input conversion module, configured to project the first error rate and the first error change rate into a second error rate and a second error change rate of the heading angle in the cloud model domain using the domain input conversion coefficients corresponding to the first error rate and the first error change rate respectively;

[0018] A first qualitative concept determination module, used to respectively determine a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve;

[0019] A second qualitative concept determination module, configured to determine a target qualitative concept corresponding to a heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept;

[0020] A domain output conversion module is used to determine the output value of the qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve, and project the output value into the output value of the basic domain according to the domain output conversion coefficient as the weight coefficient of the heading angle;

[0021] The error weight updating module is used to merge the output value with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and use the new error weight to continue to control the operation of the network intelligent robot.

[0022] In a third aspect, an embodiment of the present invention provides an electronic device, the electronic device comprising:

[0023] at least one processor; and

[0024] a memory communicatively connected to the at least one processor; wherein,

[0025] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the operation control method of the network point robot described in any embodiment of the present invention.

[0026] In a fourth aspect, a computer-readable storage medium is also provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement an operation control method of a network point robot described in any embodiment of the present invention when executed.

[0027] In a fifth aspect, a computer program product is also provided, the computer program product comprising a computer program, and the computer program, when executed by a processor, implements an operation control method of a network point robot as described in any embodiment of the embodiments of the present invention.

[0028] The technical solution of the embodiment of the present invention obtains the error rate and the error change rate by extracting the heading angle state value in real time, and determines the newly added error weight corresponding to the heading angle according to the error rate and the error change rate, and ensures that the robot constantly approaches the expected state during the movement process by continuously updating the error weight, thereby improving the overall control effect and movement performance; at the same time, the inventor creatively introduces the cloud model, and uses the combined input normal cloud curve and the combined output normal cloud curve to provide a cloud rule-based reasoning, establishes a mapping relationship between qualitative concepts and quantitative data, avoids the uncertainty and ambiguity of data in the traditional calculation process, and provides strong support for intelligent decision-making and control.

[0029] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 is a flow chart of an operation control method of a network point robot provided according to the first embodiment of the present invention;

[0032] Figure 2 is a schematic diagram of a robot attachment coordinate system and an inertial coordinate system provided according to the first embodiment of the present invention;

[0033] Figure 3 A combined input normal cloud curve reference graph provided according to the first embodiment of the present invention;

[0034] Figure 4 is a reference diagram of a heading angle coefficient adjustment rule table provided according to the first embodiment of the present invention;

[0035] Figure 5 A combined output normal cloud curve reference graph provided according to the first embodiment of the present invention;

[0036] Figure 6 is a structural schematic diagram of an operation control device of a network point robot provided according to the first embodiment of the present invention;

[0037] Figure 7 It is a structural schematic diagram of an electronic device of an operation control method of a network point robot provided according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0038] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0039] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0040] Embodiment 1

[0041] Figure 1 This is a flow chart of a method for controlling the operation of a branch robot provided by the first embodiment of the present invention. This embodiment is applicable to the case of controlling the motion of a bank branch intelligent robot. The method can be executed by a branch robot operation control device, which can be implemented in the form of hardware and / or software and can be configured in a bank branch intelligent robot. Figure 1 As shown, the method includes:

[0042] S110 . In the process of using the pre-built joint controller to control the operation of the network point robot, obtain the MPC cost function predefined in the joint controller.

[0043] The pre-structured joint controller is a controller based on the Model Predictive Control (MPC) technology, which takes into account the position coordinates, heading angle and speed of the mesh point robot. First, the above-mentioned motion state is used as a state quantity to perform model linearization to obtain a motion mathematical model. Secondly, the above-mentioned motion mathematical model is combined with the model predictive control algorithm, and a predefined MPC cost function is obtained after discretization. Since the entire control process is completely observable, the MPC cost function can be simplified to a quadratic programming problem, and the output control quantity is obtained by solving the quadratic programming problem. In the above process, the MPC cost function contains error weights. The prior art usually takes appropriate weight coefficients for the error weights based on experience, while the embodiment of the present invention extracts the heading angle state value from the original MPC cost function to update and adjust the error weight in real time. Therefore, obtaining a predefined MPC cost function is a prerequisite.

[0044] S120, extracting the heading angle state value in the mathematical model of the mesh intelligent robot motion in real time in the predefined MPC cost function, and determining the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle.

[0045] The embodiment of the present invention constructs a linear calculation formula for each state value in advance, and the motion mathematical model is a collection of multiple linear calculation formulas for state values. Therefore, the cost function obtained based on the motion mathematical model in the joint controller includes the real-time state value of each motion state of the robot.

[0046] The heading angle refers to the angle between the robot's body coordinate system x-axis and the inertial coordinate system x-axis. The body coordinate system is used to describe the relative motion of the robot, and the inertial coordinate system is the coordinate system used by the navigation system. It is easy to understand. Figure 2 This is a schematic diagram of the robot's attached coordinate system and inertial coordinate system. As shown in the figure, the inertial coordinate system is XOY, the attached coordinate system is xoy, the defined inertial coordinate system's X-axis points to the east, the Y-axis points to the north, and the defined attached coordinate system's x-axis points to the front of the robot, the y-axis points to the left of the robot, and the heading angle is Figure 2 δ f .

[0047] As the main parameter that determines the robot's motion trajectory, the error of the heading angle affects the overall control effect. Therefore, in the embodiment of the present invention, the error rate and error change rate of the heading angle can be calculated mainly by observing the heading angle state value in real time and comparing it with the preset expected heading angle value, wherein the error change rate can be obtained by derivatizing the error rate.

[0048] Since the embodiment of the present invention introduces the concept of cloud model, the first error rate refers to the heading angle error rate in the basic domain, and the first error change rate refers to the heading angle error change rate in the basic domain, that is, the actual values ​​of the first error rate and the first error change rate in real life.

[0049] Optionally, before extracting the heading angle state value in the mathematical model of the mesh intelligent robot motion in real time in the predefined MPC cost function, the following may also be included:

[0050] The coordinates, heading angle and speed of the network point intelligent robot are used as state quantities, a calculation formula for each state quantity is constructed, and the calculation formulas are combined to construct a motion mathematical model of the network point intelligent robot;

[0051] generating a joint controller according to the mathematical model of the movement of the network intelligent robot;

[0052] Reading the state quantity in the motion mathematical model to generate an error weight matrix, and constructing an MPC cost function according to the error weight matrix;

[0053] The MPC cost function is added to the joint controller.

[0054] In a specific example, the state of the robot can be expressed as (x, y, φ, v), where x and y are the current horizontal and vertical coordinates of the robot, φ is the heading angle of the robot, and v refers to the speed of the robot. A linear calculation formula is constructed for each state quantity and then combined to obtain the mathematical model of the robot's motion at a certain moment. The model predictive control (MPC) introduced in the embodiment of the present invention can transform the above optimization problem into a quadratic programming problem that is easier to solve. In this process, all known quantities in the cost function are substituted, and the retained unknown items are solved by quadratic programming.

[0055] When constructing the MPC cost function, the selection of the weight coefficient is very important. In the embodiment of the present invention, only the error weight, which is the main loss term, is considered. The error weight is composed of the above-mentioned state quantities (x, y, φ, v) and can be expressed in the form of a matrix:

[0056]

[0057] in, is the weight coefficient corresponding to the heading angle state quantity. In subsequent calculations, by continuously adjusting The value of has the effect of updating the error weight.

[0058] The embodiment of the present invention combines multiple parameters describing the robot's motion to construct a motion mathematical model that takes into account the position, direction and speed, and integrates the motion mathematical model with the model predictive control algorithm to construct an MPC joint controller. Based on the current real-time motion state of the robot, the MPC prediction, optimization and feedback characteristics are utilized to output a desired state adjustment amount.

[0059] S130. Using the domain input conversion coefficients corresponding to the first error rate and the first error change rate, respectively, the first error rate and the first error change rate are projected into the second error rate and the second error change rate of the heading angle in the cloud model domain.

[0060] The first error rate and the second error change rate are values ​​in the basic domain. Since the embodiment of the present invention introduces the concept of cloud model, it is necessary to use the domain input conversion coefficient to project the first error rate and the second error change rate in the basic domain into values ​​in the cloud model domain, that is, the second error rate and the second error rate.

[0061] Further, before projecting the first error rate and the first error change rate of the heading angle in the basic domain into the second error rate and the second error change rate of the heading angle in the cloud model domain using the domain input conversion coefficients corresponding to the first error rate and the first error change rate respectively, the method may further include:

[0062] respectively determining a first basic domain corresponding to the first error rate and a second basic domain corresponding to the first error change rate;

[0063] The ratios of the preset cloud model input domain to the first basic domain and the second basic domain are calculated respectively to obtain the first domain input conversion coefficient and the second domain input conversion coefficient.

[0064] Specifically, for the calculation of the domain input conversion coefficient, it is first necessary to determine the first basic domain according to the first error rate, and determine the second basic domain according to the first error change rate. In a specific example, if the first error rate is 15° and the first error change rate is 60° / s, the first basic domain corresponding to the first error rate is [-15°, 15°], and the second basic domain corresponding to the first error change rate is [-60° / s, 60° / s]. The cloud model input domain is a configurable item and is usually set to [-1, 1].

[0065] At this time, the ratio of the cloud model input domain to the first basic domain is calculated to obtain the first domain input conversion coefficient, and the ratio of the cloud model input domain to the second basic domain is calculated to obtain the second domain input conversion coefficient.

[0066] In an embodiment of the present invention, by determining the basic domain for the error rate and the error change rate, and combining the pre-constructed cloud model input domain, the input conversion coefficients corresponding to the error rate and the error change rate are respectively solved. This is a necessary condition for converting the error rate and the error change rate in the basic domain to the cloud model domain, and the conversion between the two domains is realized, so that the cloud model theory can be used to better serve the operation control method of a network point robot described in the present invention.

[0067] S140, respectively determining a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve.

[0068] The pre-generated combined input normal cloud curve is a curve composed of multiple normal distribution curves that meet the characteristics of the cloud model, among which NB, NS, ZE, PS and PB are qualitative concepts of the cloud model. Specifically, NB represents negative bias, NS represents negative bias, ZE represents zero, PS represents positive bias, and PB represents positive bias. The above qualitative concepts are described by a combined cloud model formed by a normal cloud and a half cloud to form a combined input normal cloud curve.

[0069] A first qualitative concept of the second error rate is determined in the combined input normal cloud curve according to its value in the cloud model domain, and a second qualitative concept of the second error change rate is determined in the combined input normal cloud curve according to its value in the cloud model.

[0070] Optionally, before respectively determining a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve according to the second error rate and the second error change rate, the method may further include:

[0071] According to the input domain of the cloud model, the input digital features of each qualitative concept are set; the digital features include expectation, entropy and hyperentropy;

[0072] Extracting the expectation in the input digital feature of each qualitative concept and taking it as the input peak value of the qualitative concept, and generating an input normal cloud curve corresponding to the current qualitative feature according to the input peak value;

[0073] The input normal cloud curves of all qualitative concepts are combined to obtain a combined input normal cloud curve whose abscissa is composed of the input expectations of the qualitative concepts.

[0074] In the cloud model, there are three digital features: expected value, entropy and super entropy. The expected value is the point value that best represents the qualitative concept in the digital space, reflecting the center of gravity of the cloud droplet group of the qualitative concept. Therefore, in the cloud model input domain, the input digital features of each qualitative concept are defined as: NB (-1, 0.17, 0.02), NS (-0.5, 0.17, 0.02), ZE (0, 0.17, 0.02), PS (0.5, 0.17, 0.02) and PB (1, 0.17, 0.02). The expected value of each qualitative concept is taken as the peak value of the normal cloud curve to quantify each qualitative concept, that is, NB is -1, NS is -0.5, ZE is 0, PS is 0.5, and PB is 1. The final combined input normal cloud curve is shown in Figure 3 , where NB and PB are semi-cloud.

[0075] In an embodiment of the present invention, a combined input normal cloud curve is obtained by constructing an input digital feature for each qualitative concept in advance, and the combined input normal cloud curve is used to realize the conversion of the error rate and the error change rate from quantitative to qualitative, which is the basis for subsequently determining the specific output value according to the target qualitative concept.

[0076] S150: Determine a target qualitative concept corresponding to the heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept.

[0077] Continuing with the above example, if the first qualitative concept is too negative (NB) and the second qualitative concept is too positive (PS), it means that the robot's course angle error is large. In this case, the heading angle weight coefficient should be appropriately increased. The heading angle is quickly restored to balance; if the first qualitative concept is close to zero (ZE) and the second qualitative concept is positively biased (NB), it means that the robot's heading angle is very small, but the heading angle is increasing rapidly. At this time, the heading angle weight coefficient should be reduced. According to the above example, a heading angle coefficient adjustment rule table for the first qualitative concept, the second qualitative concept and the target qualitative concept can be constructed, wherein the value of the target qualitative concept corresponds to The amount of adjustment.

[0078] Optionally, before determining the target qualitative concept corresponding to the heading angle in the pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept, the method may further include:

[0079] Construct a row containing all first qualitative concepts matching the heading angle in the specified order;

[0080] Construct a column containing all second qualitative concepts matching the heading angle in the specified order;

[0081] The target qualitative concept is determined according to the first qualitative concept in the row and the second qualitative concept in the column, and a heading angle coefficient adjustment rule table including all mapping relationships among the first qualitative concept, the second qualitative concept and the target qualitative concept is obtained.

[0082] According to the description of this embodiment, the heading angle coefficient adjustment rule table constructed is shown in Figure 4 ,in, refers to the second error rate corresponding to all the first qualitative concepts, Refers to the second error change rate corresponding to all the second qualitative concepts. Continuing with the above example, if the first qualitative concept is negatively biased (NB) and the second qualitative concept is positively biased (PS), the target qualitative concept is NS; if the first qualitative concept is close to zero (ZE) and the second qualitative concept is positively biased (NB), the target qualitative concept is PB.

[0083] The pre-constructed heading angle coefficient adjustment rule table lists all possible combinations of the qualitative concepts corresponding to the second error rate and the second error change rate. When any qualitative concept combination corresponding to the second error rate and the second error conversion rate is used, a clear target qualitative concept can be obtained to ensure that each adjustment of the heading angle meets actual needs.

[0084] S160, determining the value of the output qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve, and projecting the value of the output qualitative concept into the output value of the basic domain according to the domain output conversion coefficient.

[0085] When the target qualitative concept is determined, it is necessary to determine the specific value of the target qualitative concept output. Specifically, in the combined output normal cloud curve, the corresponding output qualitative concept is determined according to the target qualitative concept. At this time, the output qualitative concept is affected by the change of the output domain, and its output value is also different from the value in the combined input normal cloud curve. It should be noted that the output qualitative concept is equivalent to the qualitative concept in order to distinguish it from the qualitative concept at the time of input.

[0086] Furthermore, before determining the value of the output qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve, the method may further include:

[0087] According to the output domain of the cloud model, the output digital features of each qualitative concept are set; the digital features include expectation, entropy and hyperentropy;

[0088] Extracting the expectation in the output digital feature of each qualitative concept and taking it as the output peak value of the qualitative concept, and generating an output normal cloud curve corresponding to the current qualitative feature according to the output peak value;

[0089] The output normal cloud curves of all qualitative concepts are summarized to obtain a combined output normal cloud curve whose abscissa is composed of the output expectations of the qualitative concepts;

[0090] In the cloud model output domain, the output digital features of each qualitative concept are defined as: NB (-6, 1, 0.1), NS (-3, 1, 0.1), ZE (0, 1, 0.1), PS (3, 1, 0.1) and PB (6, 1, 0.1). The expected value of each qualitative concept is taken as the peak value of the normal cloud curve, and each qualitative concept is re-quantified, that is, NB is -6, NS is -3, ZE is 0, PS is 3, and PB is 6. The final combined output normal cloud curve is shown in Figure 5 , where NB and PB are semi-cloud.

[0091] Optionally, before projecting the output value to the output value of the basic domain according to the domain output conversion coefficient as the weight coefficient of the heading angle, the method may further include:

[0092] The ratio of the preset cloud model output domain to the third basic domain is used to obtain the domain output conversion coefficient;

[0093] The third basic domain is set based on the magnitude of the error weight matrix in the predefined MPC cost function.

[0094] Continuing with the above example, if the target qualitative concept is NS, the output value in the cloud model corresponding to the qualitative concept NS in the combined output normal cloud curve is -3. The output value in the cloud model cannot be directly input by the joint controller. It needs to be multiplied by the domain output coefficient to obtain the adjustment amount of the heading angle, and then input into the joint controller.

[0095] In an embodiment of the present invention, the cloud model output domain can be set to [-6, 6] as a configurable item, and the third basic domain is obtained by setting the magnitude of each coefficient in the error weight matrix Q, which can be [-0.05, 0.05]. The domain output conversion coefficient is then obtained by calculating the ratio. The output value of the target qualitative concept determined in the combined output normal cloud curve is multiplied by the domain output conversion coefficient to obtain the output value in the basic domain. The combined output normal cloud curve is constructed by determining the output digital features for each qualitative concept in advance, and the combined output normal cloud curve is used to realize the conversion of the heading angle weight coefficient from qualitative to quantitative, which is the basis for the subsequent update of the error weight according to the heading angle weight coefficient.

[0096] S170, combining the output value with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and using the new error weight to continue to control the operation of the network point robot.

[0097] The output value in the basic domain is compared with the original value in the error weight matrix Merge to get the new heading angle weight coefficient When the heading angle weight coefficient changes, the corresponding error weight matrix is ​​updated, that is, the original error weight matrix in the MPC cost function is updated to the new error weight. By automatically updating the error weight, the joint controller can make expected adjustments to the robot's motion state.

[0098] The embodiment of the present invention obtains the error rate and the error change rate by extracting the heading angle state value in real time, and determines the newly added error weight corresponding to the heading angle according to the error rate and the error change rate, and ensures that the robot continuously approaches the expected state during the movement process by continuously updating the error weight, thereby improving the overall control effect and movement performance; at the same time, the inventor creatively introduces the cloud model, and uses the combined input normal cloud curve and the combined output normal cloud curve to provide a cloud rule-based reasoning, establishes a mapping relationship between qualitative concepts and quantitative data, avoids the uncertainty and ambiguity of data in the traditional calculation process, and provides strong support for intelligent decision-making and control.

[0099] Embodiment 2

[0100] Figure 6This is a schematic diagram of the structure of a network point robot operation control device provided in the second embodiment of the present invention. Figure 6 As shown, the device comprises:

[0101] The cost function acquisition module 610 is used to acquire the MPC cost function predefined in the joint controller during the process of using the pre-structured joint controller to control the operation of the network intelligent robot;

[0102] The error calculation module 620 is used to extract the heading angle state value in the mathematical model of the mesh intelligent robot movement in real time in the MPC cost function, and determine the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle;

[0103] A domain input conversion module 630 is used to project the first error rate and the first error change rate into a second error rate and a second error change rate of the heading angle in the cloud model domain using the domain input conversion coefficients corresponding to the first error rate and the first error change rate respectively;

[0104] A first qualitative concept determination module 640 is used to determine the first qualitative concept corresponding to the second error rate and the second qualitative concept corresponding to the second error change rate in the pre-generated combined input normal cloud curve;

[0105] A second qualitative concept determination module 650 is used to determine a target qualitative concept corresponding to the heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept;

[0106] The domain output conversion module 660 is used to determine the output value of the qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve, and project the output value into the output value of the basic domain according to the domain output conversion coefficient;

[0107] The error weight updating module 670 is used to merge the output value with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and use the new error weight to continue to control the operation of the network intelligent robot.

[0108] The embodiment of the present invention obtains the error rate and the error change rate by extracting the heading angle state value in real time, and determines the newly added error weight corresponding to the heading angle according to the error rate and the error change rate, and ensures that the robot continuously approaches the expected state during the movement process by continuously updating the error weight, thereby improving the overall control effect and movement performance; at the same time, the inventor creatively introduces the cloud model, and uses the combined input normal cloud curve and the combined output normal cloud curve to provide a cloud rule-based reasoning, establishes a mapping relationship between qualitative concepts and quantitative data, avoids the uncertainty and ambiguity of data in the traditional calculation process, and provides strong support for intelligent decision-making and control.

[0109] Optionally, on the basis of the above embodiments, it further includes: a cost function construction unit, which is used to extract the heading angle state value in the mathematical model of the network intelligent robot movement in real time from the predefined MPC cost function, take the coordinates, heading angle and speed of the network intelligent robot as state quantities, construct a calculation formula for each state quantity, and combine the calculation formulas to obtain the mathematical model of the movement of the network intelligent robot;

[0110] generating a joint controller according to the mathematical model of the movement of the network intelligent robot;

[0111] Reading the state quantity in the motion mathematical model to generate an error weight matrix, and constructing an MPC cost function according to the error weight matrix;

[0112] The MPC cost function is added to the joint controller.

[0113] Optionally, on the basis of the above embodiments, the method further includes: an input conversion coefficient calculation unit, which is used to respectively determine a first basic domain corresponding to the first error rate and a second basic domain corresponding to the first error change rate before projecting the first error rate and the first error change rate of the heading angle in the basic domain into the second error rate and the second error change rate of the heading angle in the cloud model domain using the domain input conversion coefficients corresponding to the first error rate and the first error change rate respectively;

[0114] The ratios of the preset cloud model input domain to the first basic domain and the second basic domain are calculated respectively to obtain the first domain input conversion coefficient and the second domain input conversion coefficient.

[0115] Optionally, on the basis of the above embodiments, it further includes: an input normal cloud curve generating unit, which is used to set the input digital features of each qualitative concept according to the input domain of the cloud model before respectively determining the first qualitative concept corresponding to the second error rate and the second qualitative concept corresponding to the second error change rate in the pre-generated combined input normal cloud curve according to the second error rate and the second error change rate; wherein the digital features include expectation, entropy and super entropy;

[0116] Extracting the expectation in the input digital feature of each qualitative concept and taking it as the input peak value of the qualitative concept, and generating an input normal cloud curve corresponding to the current qualitative feature according to the input peak value;

[0117] The input normal cloud curves of all qualitative concepts are combined to obtain a combined input normal cloud curve whose abscissa is composed of the input expectations of the qualitative concepts.

[0118] Optionally, based on the above embodiments, the method further includes: an inference rule table construction unit, configured to construct a row containing all first qualitative concepts matching the heading angle arranged in a specified order before determining the target qualitative concept corresponding to the heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept;

[0119] Construct a column containing all second qualitative concepts matching the heading angle in the specified order;

[0120] The target qualitative concept is determined according to the first qualitative concept in the row and the second qualitative concept in the column, and a heading angle coefficient adjustment rule table including all mapping relationships among the first qualitative concept, the second qualitative concept and the target qualitative concept is obtained.

[0121] Optionally, based on the above embodiments, the method further includes: an output normal cloud curve generating unit, which is used to set the output digital features of each qualitative concept according to the output domain of the cloud model before determining the value of the output qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve; wherein the digital features include expectation, entropy and super entropy;

[0122] Extracting the expectation in the output digital feature of each qualitative concept and taking it as the output peak value of the qualitative concept, and generating an output normal cloud curve corresponding to the current qualitative feature according to the output peak value;

[0123] The output normal cloud curves of all qualitative concepts are summarized to obtain a combined output normal cloud curve whose abscissa is composed of the output expectations of the qualitative concepts;

[0124] Optionally, based on the above embodiments, the domain output conversion module 260 may further include: an output conversion coefficient calculation unit, which is used to calculate the ratio of the preset cloud model output domain to the third basic domain before projecting the output value into the output value of the basic domain as the weight coefficient of the heading angle according to the domain output conversion coefficient to obtain the domain output conversion coefficient;

[0125] The third basic domain is set based on the magnitude of the error weight matrix in the predefined MPC cost function.

[0126] An operation control device for a network point robot provided by an embodiment of the present invention can execute an operation control method for a network point robot provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0127] Embodiment 3

[0128] Figure 7 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0129] like Figure 7 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0130] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0131] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as an operation control method of a network point robot.

[0132] That is: in the process of using the pre-built joint controller to control the operation of the network point robot, the MPC cost function pre-defined in the joint controller is obtained;

[0133] Extract the heading angle state value in the mathematical model of the mesh intelligent robot motion in real time in the MPC cost function, and determine the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle;

[0134] Using the domain input conversion coefficients corresponding to the first error rate and the first error change rate, respectively, the first error rate and the first error change rate are projected into the second error rate and the second error change rate of the heading angle in the cloud model domain;

[0135] Determining respectively a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve;

[0136] Determining a target qualitative concept corresponding to the heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept;

[0137] Determining an output value of a qualitative concept corresponding to a target qualitative concept in a pre-generated combined output normal cloud curve, and projecting the output value into an output value of a basic domain according to a domain output conversion coefficient;

[0138] The output value is combined with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and the new error weight is used to continue to control the operation of the mesh point robot.

[0139] In some embodiments, a method for controlling the operation of a network point robot may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for controlling the operation of a network point robot described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute a method for controlling the operation of a network point robot in any other appropriate manner (e.g., by means of firmware).

[0140] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0141] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0142] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0143] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0144] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0145] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0146] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0147] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling the operation of a network point robot, characterized in that: include: In the process of using the pre-built joint controller to control the operation of the network point robot, the MPC cost function pre-defined in the joint controller is obtained; Extract the heading angle state value in the mathematical model of the mesh intelligent robot motion in real time in the MPC cost function, and determine the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle; Using the domain input conversion coefficients corresponding to the first error rate and the first error change rate, respectively, the first error rate and the first error change rate are projected into the second error rate and the second error change rate of the heading angle in the cloud model domain; Determining respectively a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve; Determining a target qualitative concept corresponding to the heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept; Determining an output value of a qualitative concept corresponding to a target qualitative concept in a pre-generated combined output normal cloud curve, and projecting the output value into an output value of a basic domain according to a domain output conversion coefficient; The output value is combined with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and the new error weight is used to continue to control the operation of the mesh point robot.

2. The method according to claim 1, characterized in that Before extracting the heading angle state value in the mathematical model of the mesh intelligent robot motion in real time in the predefined MPC cost function, it also includes: The coordinates, heading angle and speed of the network point intelligent robot are used as state quantities, a calculation formula for each state quantity is constructed, and the calculation formulas are combined to construct a motion mathematical model of the network point intelligent robot; generating a joint controller according to the mathematical model of the movement of the network intelligent robot; Reading the state quantity in the motion mathematical model to generate an error weight matrix, and constructing an MPC cost function according to the error weight matrix; The MPC cost function is added to the joint controller.

3. The method according to claim 1, characterized in that Before projecting the first error rate and the first error change rate of the heading angle in the basic domain into the second error rate and the second error change rate of the heading angle in the cloud model domain using the domain input conversion coefficients corresponding to the first error rate and the first error change rate respectively, the method further includes: respectively determining a first basic domain corresponding to the first error rate and a second basic domain corresponding to the first error change rate; The ratios of the preset cloud model input domain to the first basic domain and the second basic domain are calculated respectively to obtain the first domain input conversion coefficient and the second domain input conversion coefficient.

4. The method according to claim 1, characterized in that Before respectively determining a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve according to the second error rate and the second error change rate, the method further includes: According to the input domain of the cloud model, the input digital features of each qualitative concept are set; the digital features include expectation, entropy and hyperentropy; Extracting the expectation in the input digital feature of each qualitative concept and taking it as the input peak value of the qualitative concept, and generating an input normal cloud curve corresponding to the current qualitative feature according to the input peak value; The input normal cloud curves of all qualitative concepts are combined to obtain a combined input normal cloud curve whose abscissa is composed of the input expectations of the qualitative concepts.

5. The method according to claim 1, characterized in that Before determining the target qualitative concept corresponding to the heading angle in the pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept, the method further includes: Construct a row containing all first qualitative concepts matching the heading angle in the specified order; Construct a column containing all second qualitative concepts matching the heading angle in the specified order; The target qualitative concept is determined according to the first qualitative concept in the row and the second qualitative concept in the column, and a heading angle coefficient adjustment rule table including all mapping relationships among the first qualitative concept, the second qualitative concept and the target qualitative concept is obtained.

6. The method according to claim 1, characterized in that Before determining the value of the output qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve, it also includes: According to the output domain of the cloud model, the output digital features of each qualitative concept are set; the digital features include expectation, entropy and hyperentropy; Extracting the expectation in the output digital feature of each qualitative concept and taking it as the output peak value of the qualitative concept, and generating an output normal cloud curve corresponding to the current qualitative feature according to the output peak value; The output normal cloud curves of all qualitative concepts are summarized to obtain a combined output normal cloud curve whose abscissa is composed of the output expectations of the qualitative concepts; Before projecting the output value into the output value of the basic domain according to the domain output conversion coefficient as the weight coefficient of the heading angle, the method further includes: The ratio of the preset cloud model output domain to the third basic domain is used to obtain the domain output conversion coefficient; The third basic domain is set based on the magnitude of the error weight matrix in the predefined MPC cost function.

7. An operation control device for a network point robot, characterized in that: include: The cost function acquisition module is used to acquire the MPC cost function predefined in the joint controller during the process of using the pre-structured joint controller to control the operation of the network intelligent robot; The error calculation module is used to extract the heading angle state value in the mathematical model of the mesh intelligent robot movement in real time in the MPC cost function, and determine the first error rate and the first error change rate of the heading angle in the basic domain according to the difference between the heading angle expected value and the heading angle state angle; a domain input conversion module, configured to project the first error rate and the first error change rate into a second error rate and a second error change rate of the heading angle in the cloud model domain using the domain input conversion coefficients corresponding to the first error rate and the first error change rate respectively; A first qualitative concept determination module, used to respectively determine a first qualitative concept corresponding to the second error rate and a second qualitative concept corresponding to the second error change rate in a pre-generated combined input normal cloud curve; A second qualitative concept determination module, configured to determine a target qualitative concept corresponding to a heading angle in a pre-constructed heading angle coefficient adjustment rule table according to the first qualitative concept and the second qualitative concept; A domain output conversion module, used to determine the output value of the qualitative concept corresponding to the target qualitative concept in the pre-generated combined output normal cloud curve, and project the output value into the output value of the basic domain according to the domain output conversion coefficient; The error weight updating module combines the output value with the original heading angle weight coefficient in the error weight matrix in the MPC cost function to obtain a new error weight of the cost function, and uses the new error weight to continue to control the operation of the network intelligent robot.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute an operation control method of a network point robot as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement an operation control method of a network point robot according to any one of claims 1 to 6 when executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements an operation control method of a network point robot according to any one of claims 1-6.

Images