Abstract: According to the present invention, a method for recognition of normal and abnormal conditions can be performed with at least one neural network. First, trend data of an object system, before a recognition-step, are entered as input data to an input layer of each neural network and data of this system at the recognition-step are entered as objective output data to an output layer of the neural network. Thus, multiple sets of trend data showing at least one normal condition of this system are formed in the neural network in order to obtained learned weights and biases. Next, output data at every recognition-step are predicted by entering actual trend data as input data to the neural network, while the learned weights and biases are utilized. Then, the predicted output data are compared with actual output data at every recognition-step.

Abstract: A process for distilling multicomponent hydrocarbon mixtures in which the relative volatility between the dominant component in the lightest product and the dominant component in the heaviest component is between 1.10 and 7. The hydrocarbon mixture is fed to groups of distillation columns each of which is provided with an overhead condenser and a bottom reboiler, the columns being arranged in succession to receive the product from the prior adjacent column. The operating pressure in the columns increases from the first column to the final column within prescribed limits and the products of the mixture are separately recovered at one time in the final column without being withdrawn from intermediate distillation columns. Heat is recovered from heat source streams of the process by heat exchange with heat sink streams by bringing the streams successively in contact with a plurality of groups of heat exchangers, each group being at a different temperature level.