Testing of computer based control systems is difficult, especially because of the problem of errors in specifications. By testing software in conjunction with a model of a plant, this problem can be overcome to a large extent. By extending this idea to an abstract domain, in which tests are performed symbolically, whole classes of errors can be discovered with each test. The paper describes the basic theory for such testing.
|Place of Publication||Roskilde, Denmark|
|Publisher||Risø National Laboratory|
|Number of pages||23|
|Publication status||Published - 1981|