Search

The method of simultaneous identification and optimization aims at satisfying the conditions of optimality while providing accurate predictions of the process outputs. The model parameters are updated in a run-to-run procedure as to account for changes in operating points and to correct for errors in the predicted gradients of the cost-function and constraints. To make this parameter updating step more robust, we propose a parameter identification objective that includes a ratio of the sum of squared errors to a parametric gradient sensitivity function. This results in an identified set of parameters which provide larger sensitivities for the subsequent gradient correction step thus leading to faster convergence to the optimum. Moreover, worst-case uncertainty bounds on the model outputs are utilized to enforce an adequate model fitting. This is especially valuable when identifying dynamic metabolic models with many parameters. The resulting improvements are illustrated using two simulated cell culture processes. [ABSTRACT FROM AUTHOR]

A critical step in the control design of industrial processes is the Control Configuration Selection (CCS), where each actuator is associated with a set of measurements which will be used in the calculation of the control action. A possible solution to the CCS problem is given by the gramian-based Interaction Measures (IMs), which are derived from nominal process models. This paper introduces the derivation of uncertainty bounds for a gramian-based IM using models with uncertainty described in multiplicative form. An alternative to this model-based approach is presented, where uncertainty bounds are estimated from a tailored experiment. In addition, a procedure for robust CCS is introduced. This procedure integrates the calculated uncertainty bounds in the design of the control configuration. [ABSTRACT FROM AUTHOR]

• The sensor fault detection and isolation problems are studied for the accelerometer system in drilling tools. • Process disturbances without prior assumptions are estimated as time-varying zonotopic sets. • A residual generator with the Zonotopic Kalman Filter is proposed to deal with uncertainties, and the bounds of parametric uncertainties and measurement noises are obtained by calibration. • A fault isolation method is developed by fault directions. Sensor faults of the accelerometer system impact the efficiency and security of the drilling process, so the sensor fault detection and isolation are significant. However, there are insufficient relevant studies on the accelerometer system affected by time-varying disturbances with unknown bounds. This paper proposes a residual generator based on the Zonotopic Kalman Filter with the process disturbance estimation, and the bounds of parametric uncertainties and measurement noises are obtained by calibration. Then, a fault detection test is realized by time-varying residual bounds regarded as the combinations of residuals and dynamic thresholds. Moreover, a fault isolation decision is developed by analyzing fault directions. Finally, simulation and experiment results demonstrate the effectiveness of the proposed method. The proposed method estimates time-varying process disturbances and performs better than existing methods. This paper provides a reference for the reliable application of the accelerometer system. [ABSTRACT FROM AUTHOR]

Highlights: [•] The S101 Bridge, a prestressed concrete bridge, was damaged in a large scale progressive damage test. [•] Ambient vibration measurements during damage test provide a benchmark for monitoring algorithms. [•] Two algorithms are evaluated that take the intrinsic uncertainty in the data into account. [•] An identification and a detection approach are considered, both subspace-based. [•] The structural changes during the progressive damage test can be clearly detected. [Copyright &y& Elsevier]