Toyoda Koki Kabushiki Kaisha

Other

Masashi Yamanaka

Takao Yoneda

Kunihiko Unno

Moriaki Sakakura

Yukinori Kakazu

Shiho Hattori

Yasuji Sakakibara

Mamoru Katsuta

Tomonari Kato

Naoki Arimoto

Kazuya Hattori

Yutaka Hayashi

Masaaki Minagawa

An intelligent machining system employs a neural network for calculating machining conditions on the basis of attribute data or a workpiece and a grinding machine. The system comprises a reference machining condition calculating unit, a neural network which receives the attribute data and provides corrections, and a correcting unit for correcting the reference machining conditions by using the corrections. Corrections which cannot be determined by means of empirical expressions or theoretical expressions are determined appropriately by the neural network previously learned. The system determines corrections for the machining conditions on the basis of machining errors decided by a neural network. The system also detects time-series machining phenomena by sensors, processes the output detection signals of the sensors by a neural network to obtain machining circumstance data. The feed rate of the tool is controlled by a fuzzy inference on the basis of the machining circumstance data.

Intelligent machining system

An apparatus for carrying out learning operation of a neural network which has an input layer, an intermediate layer and an output layer. Plural nodes of the input layer are related to plural nodes of the intermediate layer with plural connection weights, while the plural nodes of the intermediate layer are also related to plural nodes of the output layer with plural connection weights. Although input data composed of plural components and teaching data composed of plural components are used in learning operation, some of the components are ineffective for the purpose of learning operation. During error calculation between output data from the neural network and the teaching data, the apparatus judges whether each of the components of the teaching data is effective or ineffective, and output errors corresponding to the ineffective components are regarded as zero. The connection weights are thereafter corrected based upon the calculated errors. The apparatus further comprises means for adding new input data and teaching data into a data base, and means for calculating a degree of heterogeneousness of the new teaching data. The new input data and teaching data are added to the data base only when the degree of heterogeneousness is smaller than a predetermined value.

Method and apparatus for performing learning in a neural network

Disclosed is a processing apparatus capable of optimizing its rule base comprising a plurality of rules having an antecedent and a consequent each. The antecedent contains input conditions of machining, and the consequent describes the result of machining. The antecedents are applied to input condition data. The consequents of the applied rules yield machining result data that are evaluated for rule fitness. When the rule base is updated, some rules are deleted and some preserved depending on their degrees of fitness. Between given two rules having the same consequent, any number of corresponding conditional elements in their antecedents are exchanged so as to generate new rules. The process is repeated until the rule base is optimized. With the antecedents of rules applied to input condition data and with the consequents of the applied rules yielding machining result data, those data may be modified according to modification request data. The consequents of rules are applied to the modified machining data, and the antecedents of these applied rules furnish corrective input condition data. This backward application of rules is an easy way to obtain corrective input condition data.

Rule fitness determination, rule removal, and new rule generation

An apparatus capable of automatically determining grinding wheel dressing timing for dressing a grinding wheel. The apparatus estimates the abrasion of the grinding wheel in a grinding cycle on the basis of machining data and grinding conditions, and divides the grinding cycle into a plurality of stages according to the estimated abrasion to carry out grinding wheel dressing after the completion of every stage of the grinding cycle. In grinding a workpiece having a plurality of ground sections, the abrasion of the grinding wheel in grinding each ground section is estimated, and the abrasion is accumulated in accordance with grinding sequence. A ground section after grinding which grinding wheel dressing is to be carried out is determined on the basis of the cumulative abrasion. Similarly, the abrasions are accumulated for each of workpieces. A grinding wheel dressing is carried out after the completion of grinding a workpiece so that the cumulative abrasion may not exceed a predetermined value.

Numerically controlled grinding machine

Growing community of inventors

Kariya, Japan

Masashi Yamanaka