JPH10289255A - Autonomous distributed design supporting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a design supporting device which conducts a design solution to design specification that is given by using an autonomous distribution technique. SOLUTION: When an autonomous element receives a stimulus from another autonomous element, a CPU updates time (S301), and next, temporarily saves just preceding parameter value and stores it to use it for calculation of a step 304 at S302. Then, it calculates all parameter values that are defined in autonomous elements at S303, reads a program for parameter evaluation processing by using a calculated parameter value and the preceding parameter value and calculates the balancing of a parameter according to the program (S304). Next, it executes autonomous evaluation processing of S305, decides the result of an autonomous element evaluating means (S306), and finishes the processing when it confirms that self-organization is accomplished.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an autonomous decentralized design support apparatus for designing mechanical products such as gears, and more particularly, to autonomous units that can be used as design elements such as parts and functions. The present invention relates to an autonomous decentralized design support device that can be defined as elements and each autonomous element performs self-organization to obtain a design solution as a whole.

[0002]

2. Description of the Related Art Conventionally, a large-scale centralized computer system has been used as a design support apparatus. The reason is that in the conventional system, the system developer describes all procedural processes, the system operates according to the description, and the system developer needs to anticipate and program all the system behavior in advance. Therefore, a large-scale centralized computer system is required to perform complicated processing.

Recently, a design support system of an autonomous decentralized type has been developed as a design support device. Here, an autonomous decentralized system does not have a control mechanism for managing the whole, and autonomous elements, which are units that can be used as design elements, cooperate and interact with each other to achieve the overall goal It is a way of thinking.

In such an autonomous decentralized system, since autonomous elements act with independence, a complicated control mechanism is not required, and high efficiency is expected in terms of reliability, expandability, flexibility and the like. .

[0005] Here, design can be interpreted as a task of obtaining a solution (design solution) by accumulating individual functions from the top down to the required specifications in a bottom-up manner. Furthermore, the process of trial and error by a designer is to set individual functions, combine them, and adjust each function while evaluating the overall goal. It is considered to be a methodology that can easily express the design process on the system.

[0006] Because of the above advantages, recently,
Autonomous decentralized systems are being developed as an alternative to large-scale, centralized computer systems.

Here, "Toshiyuki Kitamori and Shinsan Kitamura, edited by Ohmsha, Science of Self-Organization, p. 11"
According to [1], realization of self-organization is indispensable for autonomous elements to act independently. Since the self-organization is considered to generate a form and a structure as a whole by the interaction of autonomous elements, the advantage of the autonomous decentralized system changes greatly depending on how the self-organization function is realized.

For example, as a first method of self-organization,
There is a method disclosed in JP-A-7-64942 (hereinafter referred to as a first conventional method). In the first conventional method, each autonomous element is constituted by a set of simultaneous equations, and the autonomous element itself has a function of solving the simultaneous equations by itself. More specifically, it has the following functions (1) to (3).

(1) Spontaneous firing function: Fires spontaneously when necessary parameters (attributes representing the properties of the autonomous element itself) are prepared, and when the parameters are not prepared, spontaneously lacks other autonomous elements. The ability to request a value.

(2) Autonomous decision function: A function in which an autonomous element acts independently of other autonomous elements.

(3) Cooperative function: a function for responding to a request from another autonomous element.

In addition, when necessary parameters are not obtained, necessary parameters are collected by using temporary values called default values of each autonomous element, and simultaneous equations are solved.

As a second method of self-organization, there is a method disclosed in Japanese Patent Application Laid-Open No. Hei 8-137809 (hereinafter referred to as a second conventional method). This second conventional method realizes a self-organizing function by giving an appropriate time delay to a parameter calculation speed in an autonomous element and an information transmission speed between autonomous elements. This improves the property that the system tends to be clustered when the calculation speed of the autonomous element is higher than the transmission speed, and that the system tends to fall into a local solution when the calculation speed is opposite.

More specifically, each autonomous element is provided with a delay generator for transmitting the calculation result of the autonomous element after delaying it for a predetermined time, and the delay generator for all autonomous elements prepared separately is provided. It is systemized to operate according to parameters supplied from a device that handles time management collectively.

[0015]

However, the first conventional method has the following problems in a method for collecting parameter information necessary for solving the simultaneous equations.

That is, in this method, when necessary parameters are not prepared, the other autonomous element requests the missing parameter, but there is no guarantee that the other autonomous element supplies the necessary parameter value. In such a case, the default value is used, but the default value is not appropriately set to reflect the state of the autonomous element at that time, and the effect of promoting self-organization is ambiguous.

That is, it can be said that the first conventional method has a problem in that a parameter value which is not obtained by itself basically depends on another autonomous element.

Furthermore, there is a method of selectively using inquiries to other autonomous elements and using default values. However, in order to appropriately control these, it is necessary to analyze the state of the self and the state of the other autonomous elements. Control mechanism is required.

Further, in the second conventional method, the self-organization is realized by spreading a parameter calculation result in the autonomous element to another autonomous element with a time delay. The problem is that nothing is mentioned, and the autonomy is low because the time delay management mechanism does not exist in the autonomous element and depends on another management mechanism.

Here, if the autonomy becomes low, a mechanism for controlling the whole becomes necessary, and the system configuration becomes complicated and the expandability becomes low.

For the above reasons, the development of an autonomous decentralized design support apparatus capable of self-organizing autonomous elements independently of other autonomous elements with a simple system configuration has been demanded. .

The present invention has been made in view of such circumstances, and an autonomous element can function as an autonomous element having high autonomy without depending on other autonomous elements, and as a result, has a simple system configuration. It is an object of the present invention to provide an autonomous decentralized design support device capable of finding a design solution.

[0023]

According to the present invention, there is provided an autonomous decentralized design support apparatus according to the present invention based on input means, information processing means, storage means, and design specification instruction data input from the input means.
In an autonomous decentralized design support apparatus including a design means for generating a design solution by using a self-distribution technique and a display means for displaying the design solution, the design means includes an autonomous element which is a unit usable as a design element. Is a design means for calculating the design solution by self-organizing, wherein each autonomous element has a parameter as an attribute representing the property of the autonomous element itself and a constraint expression representing a relationship between the parameters, The design means includes: a parameter evaluation means for uniformly evaluating the degree of equilibrium of the parameter from the speed of the parameter value regardless of the type of the parameter; and a degree of self-organization of the autonomous element over time by the parameter evaluation means. Autonomous element evaluation means for determining based on the result; and parameter updating means for updating the parameter value based on the evaluation result of the parameter evaluation means. The design means every time receiving a stimulus from the elements is configured to calculate the design solution, the object is achieved.

Preferably, the design means is a design means for calculating the design solution according to a program for self-organizing autonomous elements, wherein the program is a program for the parameter evaluation means, Means and a program for the parameter updating means.

The operation of the present invention will be described below.

According to the above configuration, since the self-organization of the autonomous element is performed based on the change amount of the parameter of the self, the self-organization can be performed without depending on other autonomous elements. For this reason, since a complicated control mechanism is not required, the system configuration is simple.

In addition, since the degree of dependence on other autonomous elements is low, even if an autonomous element is added or changed, there is no effect on other autonomous elements, and the system can be easily expanded.

Further, unlike the above-mentioned conventional method, there is no ambiguity in the self-organization of the autonomous element, the definition of the parameter evaluation means, the autonomous element evaluation means and the parameter update means is clear, and a design support apparatus is developed. There are few points depending on the system developer. For this reason, the production of the system,
Easy maintenance and expansion.

The self-organizing process of the autonomous element, more specifically, the parameter evaluation means at each time step,
Since the designer can grasp the results of the autonomous element evaluation means and the parameter updating means, the designer can understand the progress of the self-organization with quantitative numerical values. For this reason, it is possible to construct a system having flexibility in which design specifications can be appropriately changed or added.

[0030]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be specifically described below with reference to the drawings.

FIG. 1 shows a system configuration of an autonomous distributed design support apparatus according to the present invention. This autonomous decentralized design support device,
It consists of a computer body and peripheral devices. More specifically, an information processing unit 1 (computer main body) including a CPU 1a, a ROM 1b, a RAM 1c, and an input / output circuit 1d, a keyboard 2, a mouse 3, a CRT 4 for display, and a hard disk 5 as an external storage device It is composed of

Here, the CPU 1a serves as a control center of the present system, and executes calculations described later. ROM1
b stores, in a nonvolatile manner, programs and various data necessary for the operation of the CPU 1a. The RAM 1c temporarily stores information to assist the operation of the CPU 1a. The input / output circuit 1d exchanges information between the information processing unit 1 and the peripheral device.

The keyboard 2 is used for key input of characters, numerical values, symbols, and the like, and an output result is displayed on the CRT 4. Autonomous elements and the like are stored in a hard disk 5 as an external storage device. These peripheral devices including the mouse 3 are connected to the input / output circuit 1d of the information processing unit 1, as shown in FIG.

In the above system configuration, various design support programs similar to the conventional ones are stored in the ROM 1b or the hard disk 5 in advance, and the information processing unit 1, more specifically, CP
U1a selects and starts these programs as appropriate, and performs necessary design work by inputting necessary data.

The CRT 4 appropriately stores character information, so that the designer can easily understand the process of performing the design work.
Calculation information, graphics, and the like are displayed, and information on the design result is sequentially stored in the hard disk 5.

In addition to the above design work, the autonomous decentralized design support apparatus of this embodiment has a design solution calculation function based on autonomous decentralization. Hereinafter, this function will be described.

FIG. 2 shows a configuration of a program for self-organization of autonomous elements used in the present invention, and this program is stored in the ROM 1b.

Here, the parameter 10a is the autonomous element 10
This attribute is an attribute representing the property of the keyboard 2, and is input by, for example, a designer who operates the keyboard 2. The constraint expression 10b is an expression representing the relationship between the parameters, and the parameter evaluator 10c is a parameter evaluation means, that is, a part in which a program for performing parameter evaluation processing in a program for self-organizing autonomous elements is stored. Say. Parameter update unit 1
0d denotes a parameter updating means, that is, a part in which a program for performing an autonomous element evaluation process in a program for self-organization of autonomous elements is stored.

In order to set the input specification of the designer as a parameter, for example, a flag for distinguishing between fixed and variable is added to the parameter. If a value is to be determined by organizing, a variable flag is used.

In the illustrated example, a plurality of parameters other than 10a are defined, and a parameter updating unit and a parameter evaluating unit are provided for each parameter. In addition, a plurality of constraint expressions are provided in addition to the above.

Next, the procedure of self-organization of autonomous elements will be described with reference to FIG. In the present embodiment,
Autonomous element time is discrete time (t = 0,1,2,3, ...)
You are using

The information processing unit 1, more specifically, the CPU 1a
Reads out the program for self-organization of the autonomous elements stored in the ROM 1b, temporarily stores the program in the RAM 1c, starts this processing, and when the autonomous element receives a stimulus from another autonomous element, the CPU 1a , The time is updated (step S301).

Next, in step S302, step 30
The parameter value immediately before is used for the calculation in step 4 and temporarily saved and stored in the RAM 11c. Subsequently, the CPU 1a calculates values of all parameters defined in the autonomous element 10 in step S303. Then, using the obtained parameter value and the immediately preceding parameter value, a parameter evaluation unit stored in the parameter evaluation unit 10c, that is, a program for parameter evaluation processing is read out, and the equilibrium degree of the parameter is obtained according to this program ( Step S304).

Here, the parameter evaluation means used in step S304 standardizes the difference between the parameter values of the immediately preceding time and the current time, and is expressed by the following equation (1).

[0045]

(Equation 1)

In the equation (1), the denominator ma on the right side is
x is a function that returns the larger value of the given arguments. X indicates a parameter 10a set in the autonomous element 10. Incidentally, in a gear design example described later, each gear is defined as the autonomous element 10, and items defining characteristics of the autonomous element 10, such as the number of teeth, the number of revolutions, and the pitch circle diameter, are used as the parameters 10a.

Here, since the parameter 10a changes with time, it is generally represented by a function x (t) of time.
However, since it is a discrete time, it is expressed as t = 0, 1, 2,.

The above equation (1) indicates that the parameter 10a is closer to equilibrium when ρ is closer to 1, and that the parameter 10a is not balanced when ρ is closer to 0.

When t = 0, there is no value to be compared, so an appropriate value is set and ρ = 0.5. However, when the parameter value is obtained, the constraint equation involving the parameter 10a is evaluated, and when one of the constraint equations is not satisfied, ρ = 0.

This updating process is performed in the following procedure.

(1) Sort the current equilibrium in the parameters of each autonomous element in descending order. Then, the following processes (2) to (4) are repeated in ascending order of the parameter values.

(2) If the other parameter values necessary for obtaining the parameter value are already set in the autonomous element or can be obtained from the other autonomous element, solve the necessary relational expression. Calculate the parameter value.

(3) If the information for determining the target parameter value is insufficient and it is not possible to obtain the parameter value by solving the relational expression, the following expressions (2) and (3) are applied.

X (t + 1) = x (t) ± (1-ρ (t)) × STEP (2) x (t + 1) = x (t) ± α × STEP (3) Equations (2) and (3) show a kind of time-delayed feedback control, in which parameter values reflecting the state at the immediately preceding time are set. Note that the notation STEP in the equations (2) and (3) indicates the step size of the parameter uniquely set for each parameter value.

The above equation (3) shows a case where sufficient clues for updating the parameter are not obtained, and a random number is generated within the range of the step size in the manner of random search to determine the value.
Α in the expression (3) represents a random number of 0 <α <1.

(4) As a method of updating the parameter value, there are two ways of increasing and decreasing the parameter value. Therefore, in each case, the parameters are updated and evaluated, and a method in which the degree of balance of the autonomous element is improved as compared with the immediately preceding value is adopted. If neither of them is improved, the parameter is not updated.

When the processing in step S304 is completed, C
The PU 1a executes the autonomous evaluation processing of Step S305. In this autonomous evaluation processing, the degree of balance of the autonomous elements is obtained by using the autonomous evaluation means stored in the autonomous element evaluation unit 10e. This autonomous evaluation means is an average value of the state quantities of the parameters represented by the above equation (1) and is represented by the following equation (4).

[0058]

(Equation 2)

In the above equation (4), the value of the average value ψ is 1
The closer to, the higher the self-organization degree of the autonomous element, and the closer to 0, the lower the self-organization degree.
Here, n indicates the number of parameters defined in the autonomous element.

Next, the CPU 1a determines the result of the autonomous element evaluation means (step S306), and confirms that the self-organization has been achieved, at which point the processing for self-organization of the autonomous element ends. I do. On the other hand, if it is determined that self-organization has not been achieved, the process returns to step 301 and waits for a new stimulus. If there is a new stimulus, the processes of steps S301 to S306 are repeated.

Next, the above-mentioned method will be specifically described by taking the design of a gear train as an example. The design target is a case where a specified number of spur gear trains are arranged in a rectangle on a two-dimensional plane.

In this case, the autonomous decentralized design support apparatus handles two autonomous elements, ie, an element representing a rectangular outer shape and a gear element.

Here, the autonomous element representing the outer shape is, for example,
It has four parameters of a minimum value X _min of X (X coordinate), a maximum value X _max and a minimum value Y _min of Y (Y coordinate) and a maximum value Y _max . The autonomous elements of the gear include, as parameters, the number of teeth (X ₁ ), module (X ₂ ), pitch circle diameter (X ₃ ),
Six parameters of the rotation speed (X ₄ ), the X coordinate (X ₅ ) and the Y coordinate (X ₆ ) of the gear center, and the following equation (5.1)
It is composed of nine relational expressions (constraint expressions) represented by the expression (5.9).

Although the above parameters are not all elements that determine the gear, typical parameters are listed for describing the embodiment of the present invention.

[0065] _{14 ≦ X 1 ... (5.1)} X 2 is according to JIS B1701 module first sequence _{... (5.2) X 5 -X min} > (X 3/2) + X 2 ... (5.3) X _{max -X 5> (X 3/} 2) + X 2 ... (5.4) X 6 -Y min> (X 3/2) + X 2 ... (5.5) Y max -X 6> (X 3/2 ) + X ₂ (5.6) X ₃ = X ₁ × X ₂ (5.7) X ₄₁ / X ₄₂ = X ₁₂ / X ₁₁ (5.8) d = X ₃₁ + X _32/2 ( 5.9) Here, the above equation (5.1) is an equation representing the limit number of teeth of the gear. The above equation (5.2) is an equation representing a gear module. Equations (5.3) to (5.6) above are equations representing interference between a gear and a rectangle. The above equation (5.7) is an equation representing the relationship between the number of teeth, module and pitch circle diameter. The above equation (5.8) is an equation representing the relationship between the number of rotations and the number of teeth. The above equation (5.9) is an equation representing the distance between the gear centers.

Since the above equations (5.8) and (5.9) are relational equations between the two gears, the subscript 2
The second number represents the gear number.

In the case of a gear, as long as only a drive gear and a driven gear can be defined, the rest are all intermediate gears, and the autonomous elements can interact with each other to easily define an adjacent relationship with each other.

As a means for giving a stimulus to each autonomous element, for example, one token is prepared and given to an arbitrary autonomous element.
The autonomous element that receives the token interprets it as a stimulus,
react. Then, the operation of transferring the token to the adjacent gear is propagated one after another.

If it is considered that the driving gear and the driven gear are virtually adjacent to each other, a periodic loop is formed, so that the calculation can be repeated until all the autonomous elements have completed self-organization.

FIG. 4 shows the initial state of the spur gear train when the rotational speeds and the center coordinates of the driving gear and the driven gear are given, and FIG. 5 shows the spur gear train after self-organization. That is, it shows the state of the spur gear train in the design solution obtained by the above method.

[0071]

According to the autonomous distributed design support apparatus of the present invention, since the self-organization of the autonomous elements is performed based on the amount of change of the parameters owned by the self, the self-organization of the autonomous elements is independent of other autonomous elements. Organization becomes possible. For this reason, since a complicated control mechanism is not required, the system configuration is simple. Therefore, an inexpensive autonomous decentralized design support device can be realized.

In addition, since the degree of dependence on other autonomous elements is low, even if an autonomous element is added or changed, there is no effect on other autonomous elements, and the system can be easily expanded.

Further, unlike the above-mentioned conventional method, there is no ambiguity in the self-organization of the autonomous element, the definition of the parameter evaluation means, the autonomous element evaluation means and the parameter update means is clear, and a design support apparatus is developed. There are few points depending on the system developer. For this reason, the production of the system,
Easy maintenance and expansion.

The self-organizing process of the autonomous element, more specifically, the parameter evaluation means at each time step,
Since the designer can grasp the results of the autonomous element evaluation means and the parameter updating means, the designer can understand the progress of the self-organization with quantitative numerical values. For this reason, it is possible to construct a system having flexibility in which design specifications can be appropriately changed or added.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a system configuration of an autonomous distributed design support apparatus of the present invention.

FIG. 2 is a diagram showing a configuration of a program for self-organization of autonomous elements used in the present invention.

FIG. 3 is a flowchart showing a procedure of self-organization of autonomous elements.

FIG. 4 is a diagram showing a state of a spur gear train before self-organization when the present invention is applied to a spur gear train design.

FIG. 5 is a diagram showing a state of a spur gear train in a design solution obtained by executing self-organization from the state of FIG. 4;

[Explanation of symbols]

 Reference Signs List 1 information processing unit 1a CPU 1b ROM 1c RAM 1d input / output circuit 2 keyboard 3 mouse 4 CRT 5 hard disk 10 autonomous element 10a parameter 10b constraint expression 10c parameter evaluation unit 10d parameter update unit 10e autonomous element evaluation unit

Claims (2)

[Claims] An input unit, an information processing unit, a storage unit, a design unit for generating a design solution by using a self-distribution method based on design specification instruction data input from the input unit, and displaying the design solution. In an autonomous distributed design support apparatus including a display unit, the design unit is a design unit that calculates the design solution by self-organizing each autonomous element that is a unit usable as a design element, The autonomous element has a parameter as an attribute representing the property of the autonomous element itself and a constraint expression representing a relation between the parameters. The design means determines the degree of equilibrium of the parameter from the speed of the parameter value according to the type of the parameter. Parameter evaluation means for unified evaluation, and autonomous element evaluation for determining the degree of self-organization of the autonomous element over time based on the evaluation result of the parameter evaluation means An autonomous decentralized design, comprising: a stage; and a parameter updating unit that updates the parameter value based on the evaluation result of the parameter evaluation unit. Support equipment. 2. The design means for calculating the design solution according to a program for self-organizing autonomous elements, the program comprising: a program for the parameter evaluation means; and a program for the autonomous element evaluation means. 2. The autonomous decentralized design support apparatus according to claim 1, further comprising a program for the parameter updating means and a program for the parameter updating means.