US 6,148,274 A · Optimization adjusting method and optimization adjusting apparatus

A computer method for estimating a solution by successively improving solution vectors, the solution vectors being set for a problem having time sequential information, said computer method comprising the steps of:

(a) inputting time sequential information associated with the operation of a target apparatus;
(b) setting an initial set of solution vectors;
(c) presenting said time sequential information expressed by each solution vector to a user;
(d) receiving a value from the user which is indicative of an evaluation by the user of the suitability for solving the problem of each solution vector;
(e) rearranging in accordance with each value received from the user the order in which the tine sequential information expressed by each solution vector are presented to the user;
(f) calculating a fitness value of each solution vector based on the value received from the user;
(g) performing an arithmetical genetic recombination operation on the solution vectors in the solution vector set based on the fitness value of each solution vector, thereby constructing a new solution vector set;
and
(h) adjusting the target apparatus operating on the time sequential information inputted in step (a) based on the solution,
wherein step (c) through step (g) are iteratively performed until one of a number of iterations of steps (c) through (g) and the fitness value of one solution vector exceeds a predetermined threshold, thereby estimating the solution for the user.

7 more elements

A computer method for estimating a solution by successively improving solution vectors, the solution vectors being set for a problem having time sequential information, said computer method comprising the steps of:

(a) inputting time sequential information associated with the operation of a target apparatus;
(b) setting an initial set of solution vectors;
(c) presenting said time sequential information expressed by each solution vector to a user;
(d) receiving a value from the user which is indicative of an evaluation by the user of the suitability for solving the problem of each solution vector;
(e) rearranging in accordance with each value received from the user the order in which the time sequential information expressed by each solution vector are presented to the user;
(f) recording the the respective value received from the user and the respective solution vector as a history;
(g) determining whether the history is consistent with a predetermined evaluation model estimation condition;
(h) receiving a fitness value associated with each solution vector from the user when it is determined in step (g) that the history is not consistent with the predetermined evaluation model estimation condition;
(i) performing an arithmetical genetic recombination operation on the solution vectors in the solution vector set based on the fitness value of each solution vector, thereby constructing a new solution vector set;
(j) calculating an evaluation model for the user based on the history when it is determined in step (g) that the history is consistent with the predetermined evaluation model estimation condition;
(k) outputting the solution based on the evaluation model;
and
(l) adjusting the target apparatus operating on the time sequential information inputted in step (a) based on the solution outputted in step (k),
wherein step (b) through step (i) are iteratively performed until the history is consistent with the predetermined evaluation model estimation condition.

11 more elements

A computer method for estimating a solution by successively improving solution vectors, the solution vectors being set for a problem having time sequential information, said computer method comprising the steps of:

(a) inputting time sequential information associated with the operation of a target apparatus;
(b) setting an initial set of solution vectors;
(c) presenting said time sequential information expressed by each solution vector to a user;
(d) receiving a value from the user which is indicative of an evaluation by the user of the suitability for solving the problem of each solution vector;
(e) rearranging in accordance with each value received from the user the order in which the time sequential information expressed by each solution vector are presented to the user;
(f) recording the the respective value received from the user and the respective solution vector as a history;
(g) calculating a fitness value associated with each solution vector based on the value received from the user;
(h) performing an arithmetical genetic recombination operation on the solution vectors in the solution vector set based on the fitness value of each solution vector, thereby constructing a new solution vector set;
(i) when a predetermined user evaluation termination condition is satisfied, determining whether a predetermined number of users satisfy a predetermined common model estimation condition;
(j) when it is determined in step (i) that the predetermined number of users satisfy the predetermined user evaluation termination condition, calculating a common adjustment process model based on the history of the predetermined number of users;
(k) outputting the solution based on the common adjustment process model;
and
(l) adjusting the target apparatus operating on the time sequential information inputted in step (a) based on the solution outputted in step (k),
wherein steps (c) through (i) are preformed iteratively until the predetermined user evaluation termination condition is satisfied.

11 more elements

A computer method for estimating a solution by successively improving solution vectors, the solution vectors being set for a problem having time sequential information, said computer method comprising the steps of:

(a) inputting time sequential information associated with the operation of a target apparatus;
(b) setting an initial set of solution vectors;
(c) assigning each solution vector a value based on a previously extracted common model representing a plurality of users' adjustment processes;
(d) presenting said time sequential information expressed by each solution vector to a user;
(e) receiving an indication from the user whether the value assigned to each solution vector in step (c) using the common model is representative of an evaluation by the user of the suitability for solving the problem of each solution vector;
(f) receiving the value from the user which is representative of the evaluation by the user of the suitability for solving the problem of each solution vector if the indication received from the user in step (e) is that the value assigned to each solution vector in step (c) using the common model is not representative of the evaluation by the user;
(g) rearranging in accordance with each value assigned in step (c) and received in step (f) the order in which the time sequential information expressed by each solution vector arc presented to the user;
(h) if the value is received from the user in step (f), recording the respective solution vector and the respective value received for the user as a history;
(i) determining whether the history is consistent with a common model update condition;
(j) if it is determined in step (i) that the history is consistent with the common model update condition, updating the common model representing the plurality of users' adjustment processes by using the history recorded in step (h);
(k) receiving a fitness value associated with each solution vector from one of the value received from the user and the value assigned to each solution vector based on the updated common model;
(l) performing an arithmetical genetic recombination operation on the solution vectors in the solution vector set based on the fitness value of each solution vector, thereby constructing a new solution vector set;
(m) outputting the solution;
and
(n) adjusting the target apparatus operating on the time sequential information inputted in step (a) based on the solution outputted in step (m),
wherein by iteratively performing step (c) through step until one of a number of iterations of steps (c) through (l) and the fitness value of one solution vector exceeds a predetermined threshold.

13 more elements

A computer method for estimating a solution by successively improving solution vectors, the solution vectors being set for a problem having time sequential information, comprising:

(a) inputting time sequential information associated with the operation of a target apparatus;
(b) setting an initial set of solution vectors;
(c) presenting said time sequential information expressed by each solution vector to a user;
(d) receiving a value form the user which is indicative of an evaluation by the user of the suitability for solving the problem of each solution vector;
(e) rearranging in accordance with each value received from the user the order in which the time sequential information expressed by each solution vector are presented to the user;
(f) recording the respective value received from the user and the respective solution vector as a history;
(g) determining whether the history is consistent with a predetermined evaluation model estimation condition;
(h) calculating an evaluation model based on the history, when it is determined in step (g) that the history is consistent with the predetermined evaluation model estimation condition;
(i) calculating a fitness value for each solution vector based on the value received from the user;
(j) performing an arithmetical genetic recombination operation on the solution vectors in the solution vector set based on the fitness value of each solution vector, thereby constructing a new solution vector set;
(k) outputting the solution based on the evaluation model;
and
(l) adjusting the target apparatus operating on the time sequential information inputted in step (a) based on the solution outputted in step (k),
wherein the evaluation model is calculated based on the history recorded by iteratively performing steps (c) through (j) until the evaluation model estimation condition is consistent with the history, and thereafter, iteratively performing steps (h) through (j) until a predetermined termination condition is satisfied.

11 more elements