Abstract: The present invention discloses a method of reducing octane loss in catalytic cracking of gasoline in S-Zorb plant. The method comprises following steps: a. Collecting equipment data and adjusting differences thereamong to construct initial data sample set; b. Collating data sample set; c. Employing local linear embedding to reduce dimension of independent variables, and trimming the independent variables to obtain reserved independent variables; d. Calculating correlation coefficients of the reserved independent variables and removing variables with too high correlation coefficients to obtain main variables; e. Constructing a BP neural network model of the main variables to obtain optimal octane loss prediction model; f. Employing the optimal octane loss prediction model to obtain a predicted optimal operating condition for each main variable. The present invention has advantage of reducing octane loss and improving quality of refined oil.

Abstract: The invention provides a group of anti-VHH domain antibodies and use thereof. The present invention further provides use of the described antibodies in the development, screening and purification of nano-antibodies, and use of the described antibodies in the field of immunotherapy.

Abstract: The present invention discloses a method of reducing octane loss in catalytic cracking of gasoline in S-Zorb plant. The method comprises following steps: a. Collecting equipment data and adjusting differences thereamong to construct initial data sample set; b. Collating data sample set; c. Employing local linear s embedding to reduce dimension of independent variables, and trimming the independent variables to obtain reserved independent variables; d. Calculating correlation coefficients of the reserved independent variables and removing variables with too high correlation coefficients to obtain main variables; e. Constructing a BP neural network model of the main variables to obtain optimal octane loss prediction model; f. Employing the optimal octane loss prediction model to obtain a predicted optimal operating condition for each main variable. The present invention is different from traditional chemical mechanism method at aspect of building FCC technology process.