Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – With indicator or control of power plant
Reexamination Certificate
1999-07-02
2004-04-13
Dolinar, Andrew M. (Department: 3747)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
With indicator or control of power plant
C700S048000, C706S013000
Reexamination Certificate
active
06721647
ABSTRACT:
BACKGROUND
1. Field of the Invention
The present invention relates to a method for evaluating a genetic algorithm, and particularly to a genetic algorithm used in a control system for an engine.
2. Related Art
When a control system or control characteristics of a machine, such as a vehicle or electrical appliance, is designed, imaginary users are selected, and the users' preferences and their using conditions are taken into consideration. The characteristics of the controlled machine (plant) are determined in such a way as to adapt the machine to users in as broad a range as possible.
However, each individual user has a particular and unique personality, and their preferences are diverse. Thus, there is a problem in that even if imaginary users are selected to develop and design a product for the users by presuming the users' preference, it is impossible to satisfy all of the users of the product.
In order to solve the above problem, prior to purchase of a product, in some cases a prospective user is requested to determine whether or not the product is satisfactory to the user after checking the characteristics of the product in light of the user's preferences. However, it is troublesome for the user to check the characteristics of the product before the purchase. Further, because a series of products are often operated or controlled by characteristics common in the products, although the design of the product is changed depending on the user's preferences, the user may not like other operational characteristics. Thus, although the design is appealing to some prospective users, the users may not purchase the product since the operational characteristics do not appeal to them. In the other words, there is another problem in that the range of users is limited and depends on the operational characteristics.
Online (i.e. during use by the user) adaptation of the control system using a genetic analyzer to accommodate user preferences is difficult because 1) the online system may not provide enough data for the genetic analyzer to operate properly (i.e. the online system may operate with a reduced sensor set) and 2) because the online genetic analyzer must operate very quickly in real, or near-real, time.
SUMMARY OF THE INVENTION
The present invention solves these and other problems by providing a way to estimate the best children (i.e. the next generation of individual chromosomes) for use by an online genetic analyzer (GA). One embodiment provides a method for evaluation of a GA in a control system for controlling a plant wherein control outputs appear in multiple operational stages of the controlled plant. Individuals are evaluated using data for evaluation. The data for evaluation is categorized into subdomains of a larger evaluating domain. In one embodiment, the data for evaluation is sorted into an evaluation area wherein an evaluation value of each individual is replaced by a model of coefficients relative to a standard model. In one embodiment, the standard model is used to calculate an evaluation value. The evaluation value can include a mean value and/or a distribution. Moreover, the mean value and distribution can be valid for an area including a periphery of the subdomain. The evaluation value can be calculated as a model obtained by using a least-squares method to find coefficients relative to coefficients of the standard model. Any controlled plant can be used, including engines, electric motors, air-conditioners, refrigerators, robots, etc.
Parameters for subdividing the evaluation domain can include items such as user preference, user skill level, plant operating conditions, environmental states, and/or other operating states (e.g. transient states, steady state, etc).
When used with an internal combustion engine, the optimized parameters can include fuel efficiency and response performance (e.g. power/acceleration).
One embodiment includes an integrated control method comprising the steps of: determining the characteristics of a user and/or using conditions; and changing characteristics of a control system of a machine in accordance with the determined characteristics. In the above, the machine is operable by a causative signal, and preferably, the control system first outputs a base value of the causative signal to control the machine based on indicative signals indicating a result of the control of the machine, and the changing step including the steps of: creating multiple control modules for representing at least one factor to be controlled; selecting at least one control module most adaptable for a current operational state based on the determined characteristics of the user and/or using conditions; learning information from a control module; compensating for the base value based on the result of the selection and the learning; and controlling the machine using the output compensated for.
According to the present invention, the machine is “trained” to suit the characteristics of the user and/or the using conditions, thereby easing control of the machine particularly for the user and enjoying training and adapting the machine to the user's preference.
In one embodiment, the control system includes: a reflection hierarchy for outputting the base value reflectively in response to input from the using conditions; an evolutionary-adaptation hierarchy for conducting the creating step, the selecting step, and the compensating step; and a learning hierarchy for conducting the learning step and the compensating step. In one embodiment, the learning hierarchy includes a control system for learning and a control system for operation, both control systems being interchangeable. While the control system for learning is learning, the control system for operation is controlling the machine in cooperation with the reflection hierarchy.
In the above, preferably, the evolutionary-adaptation hierarchy is inactivated when the control system for learning completes learning. Further, after being inactivated, the evolutionary-adaptation hierarchy is activated at given intervals to check drift between an actual state and a state controlled by the reflection hierarchy and the control system for operation in the learning hierarchy, and when there is drift, the evolutionary-adaptation hierarchy resumes the creating step and the selecting step. Accordingly, by checking the control particulars at given intervals, it is possible to constantly maintain the most suitable operation against a change in the using environment or deterioration with age.
Further, in the above method, parameter-obtaining devices are not newly required. Existing devices can be used for obtaining necessary parameters, thereby lowering the cost.
When the machine to be controlled is an engine installed in a vehicle, the operation characteristics of the engine can be changed to suit the driver's preferences, and when control is conducted based on the driver's skill, suitable driving performance can be realized in accordance with the driver's skill and its improvement.
Since the user can train the machine (engine) based on the user's preference after its purchase, the user can give less weight to the characteristics of the engine itself, and can select a vehicle from a wide range at purchase.
When the machine to be controlled is an auxiliary power unit installed in a bicycle or a wheelchair, the characteristics of the auxiliary power unit (motor) can be changed to suit the user's preferences, thereby effecting assistance most customized to each individual user.
When the machine to be controlled is a robot, the characteristics of the robot can be changed to suit the user's preferences, thereby operating the robot in a way most suitable to each individual user.
When the machine to be controlled is a suspension system or seat, the characteristics of the damper of the suspension system or seat can be changed to suit the user's preferences, thereby obtaining characteristics of the damper most suitable to each individual user.
When the machine to be controlled is a s
Kita Hajime
Mizutani Takaaki
Yamaguchi Masashi
Dolinar Andrew M.
Knobbe Martens Olson & Bear LLP
Yamaha Hatsudoki Kabushiki Kaisha
LandOfFree
Method for evaluation of a genetic algorithm does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for evaluation of a genetic algorithm, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for evaluation of a genetic algorithm will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3250101