Parameter optimization method, device, and storage medium
Examples of the present disclosure provide a parameter optimization method, device and computer readable storage medium. The method includes: establishing a one-to-one functional relationship between each parameter and each performance index; determining each current correlation coefficient between the parameter and each performance index based on the one-to-one function relationships; obtaining a current weight of each performance index; according to current weights and current correlation coefficients, obtaining a current influence coefficient of each parameter on a comprehensive performance of the performance indexes; and determining important optimization parameters according to the current influence coefficient; for each two parameters, calculating a current correlation coefficient of the two parameters, and determining an adjustment parameter; and performing parameter optimization based on the important optimization parameters and the adjustment parameters. The technical solutions of the present disclosure can improve the parameter optimization efficiency.
1 . A parameter optimization method comprising:
establishing a one-to-one functional relationship between each parameter of a set of parameters and each performance index of a set of performance indexes based on function relationships between each performance index and all parameters of the set of parameters;
determining each current correlation coefficient between the parameter and each performance index based on the one-to-one function relationships;
obtaining a current weight of each performance index;
according to the current weights and the current correlation coefficients, obtaining a current influence coefficient of each parameter on a comprehensive performance of the performance indexes;
determining a parameter whose current influence coefficient reaches a set high threshold as an important optimization parameter, or determining a first one of the one or more parameters with a highest current influence coefficient of the one or more parameters as the important optimization parameter;
for each two parameters, calculating a current correlation coefficient of the two parameters according to the current weights and the current correlation coefficient between each of the two parameters and each performance index, and determining the parameter whose current correlation coefficient with the important optimization parameter meets set requirements as an adjustment parameter;
performing parameter optimization based on the important optimization parameters and the adjustment parameters;
establishing a knowledge map of the parameters based on an association coefficient of each of the two parameters;
wherein the knowledge map comprises nodes representing the parameters and connecting lines between nodes representing an association relationship between parameters; and
in the knowledge map, the size of each node is directly proportional to the value of the influence coefficient of the parameter represented by the node, and the nodes corresponding to the important optimization parameters and the adjustment parameters are highlighted.
2 . The method according to claim 1 , wherein determining each current correlation coefficient between the parameter and each performance index based on the one-to-one function relationships comprises:
for a current value of each parameter, determining a tangent slope of the current value in a function image of the one-to-one functional relationship between the parameter and a performance index as the current correlation coefficient between the parameter and the performance index.
3 . The method according to claim 1 , further comprising establishing a correlation coefficient table between each parameter and each performance index utilizing the current correlation coefficient between each parameter and each performance index.
4 . The method according to claim 3 , further comprising adopting different colors in the correlation coefficient table to represent different values of the current correlation coefficients to obtain a correlation nephogram of the parameters and the performance indexes.
5 . The method according to claim 3 , wherein obtaining the current influence coefficient of each parameter on the comprehensive performance of the performance indexes comprises:
bringing current weight of each performance index into the correlation coefficient table to obtain a weighted correlation coefficient table; and
adding a weighted current correlation coefficient corresponding to the same parameter in the weighted correlation coefficient table to obtain the current influence coefficient of each parameter on the comprehensive performance.
6 . The method according to claim 1 , further comprising:
calculating a comprehensive score representing the comprehensive performance of the performance indexes according to the weights of the performance indexes and the current values of the performance indexes;
when the comprehensive score is greater than or equal to a predetermined threshold, determining parameters corresponding to the comprehensive score as final optimization parameters; otherwise, performing the process of determining each current correlation coefficient between the parameter and each performance index based on the one-to-one function relationships;
wherein performing parameter optimization based on the important optimization parameters and the adjustment parameters comprises:
adjusting values of the parameters according to a predetermined rule including adjusting values of the important optimization parameter and adjustment parameters first; or adjusting only values of the important optimization parameter and adjustment parameters; then
returning to the process of establishing the one-to-one functional relationship between each parameter and each performance index based on the function relationships between each performance index and all parameters.
7 . The method according to claim 6 , further comprising, in response of a change of the current weights, returning to the process of calculating a comprehensive score representing a comprehensive performance of the performance indexes according to the weights of the performance indexes and the current values of the performance indexes.
8 . The method according to claim 6 , further comprising, in response to a change of the important optimization parameter and adjustment parameters, performing the parameter optimization based on changed important optimization parameters and the adjustment parameters.
9 . A parameter optimization device comprising:
a first hardware module operating to establish a one-to-one functional relationship between each parameter and each performance index based on function relationships between each performance index and all of the parameters;
a second hardware module operating to determine each current correlation coefficient between the parameter and each performance index based on the one-to-one function relationships;
a third hardware module operating to obtain a current weight of each performance index;
a fourth hardware module operating to obtain a current influence coefficient of each parameter on a comprehensive performance of the performance indexes according to the current weights and the current correlation coefficients, and to determine a parameter whose current influence coefficient reaches a set high threshold as an important optimization parameter, or to determine a first of the one or more parameters as the parameter with a highest current influence coefficient of the one or more parameters as an important optimization parameter;
a fifth hardware module, for each two parameters, operating to calculate a current correlation coefficient of the two parameters according to the current weights and the current correlation coefficient between each of the two parameters and each performance index, and to determine the parameter whose current correlation coefficient with the important optimization parameter meets set requirements as an adjustment parameter; and
a sixth hardware module operating to perform parameter optimization based on the important optimization parameters and the adjustment parameters.
10 . The device according to claim 9 , further comprising:
a seventh module operating to establish a correlation coefficient table between each parameter and each performance index utilizing the current correlation coefficient between each parameter and each performance index;
in the correlation coefficient table, different colors are adopted to represent different values of the current correlation coefficients to obtain a correlation nephogram of the parameters and the performance indexes.
11 . The device according to claim 10 , wherein the fourth module operates to bring the current weight of each performance index into the correlation coefficient table to obtain a weighted correlation coefficient table, and to add a weighted current correlation coefficient corresponding to the same parameter in the weighted correlation coefficient table to obtain the current influence coefficient of each parameter on the comprehensive performance.
12 . The device according to claim 9 , further comprising:
an eighth module operating to establish a knowledge map of parameters based on an association coefficient of each of the two parameters;
wherein the knowledge map comprises nodes representing the parameters and connecting lines between nodes representing an association relationship between the parameters; and
in the knowledge map, the size of each node is directly proportional to the value of the influence coefficient of the parameter represented by the node, and the nodes corresponding to the important optimization parameters and the adjustment parameters are highlighted.
13 . The device according to claim 9 , further comprising:
a ninth module operating to calculate a comprehensive score representing the comprehensive performance of the performance indexes according to the weights of the performance indexes and the current values of the performance indexes; and
a tenth module operating to determine parameters corresponding to the comprehensive score as final optimization parameters when the comprehensive score is greater than or equal to a predetermined threshold; otherwise, to inform the second module to perform corresponding processes;
wherein the sixth module further operates to adjust values of the parameters according to a predetermined rule including adjusting values of the important optimization parameter and adjustment parameters first, or to adjust only values of the important optimization parameter and the adjustment parameters, then indicate the first module to perform corresponding processes.
14 . A parameter optimization device comprising:
at least one non-transitory memory storing a computer program; and
at least one processor to call the computer program stored in the at least one non-transitory memory to perform a parameter optimization method including:
establishing a one-to-one functional relationship between each parameter of a set of parameters and each performance index of a set of performance indexes based on function relationships between each performance index and all parameters of the set of parameters;
determining each current correlation coefficient between the parameter and each performance index based on the one-to-one function relationships;
obtaining a current weight of each performance index;
according to the current weights and the current correlation coefficients, obtaining a current influence coefficient of each parameter on a comprehensive performance of the performance indexes;
determining a parameter whose current influence coefficient reaches a set high threshold as an important optimization parameter, or determining a first of the one or more parameters with a highest current influence coefficient among the one or more parameters as an important optimization parameter;
for each two parameters, calculating a current correlation coefficient of the two parameters according to the current weights and the current correlation coefficient between each of the two parameters and each performance index, and determining the parameter whose current correlation coefficient with the important optimization parameter meets set requirements as an adjustment parameter; and
performing parameter optimization based on the important optimization parameters and the adjustment parameters.