SD2F2 : FEEDFORWARD PROCESS CONTROL: an mws model,
Dennis Ridley, Ph.D. Florida A&M University, Tallahassee, Fl 32307 and Supercomputer Computations Research Institute, Florida State University, Tallahassee, Fl 32306,
and
Felipe Llaugel, M.Eng., Science and Technology Center, Universidad Dominicana O&M.

Feedforward information about a measurable disturbance is proactively fed, ahead of time, to the manipulated inputs of a process, the output of which is to be controlled, so as to counteract the effect of the disturbance. Discretized observations on the process variable are indexed to form a time series, to which a model is fitted. Ultra high signal to noise ratio fitted values are examined by a neural network, for common cause patterns which detect when the future process will become out of control. Corrective action is initiated to prevent the process from ever actually being out of control.

INTRODUCTION

The purpose of this paper is to present a feedforward control procedure which is designed to prevent a process from ever being out of control. If the assumption of an independently and identically distributed (IID) process variable were true, then clearly this would be impossible since there would be no way to predict an out of control condition. It turns out however, that in actual manufacturing and other processing systems, the process variables are not IID. These systems exhibit systematic serial correlation estimated to be as high as 80% ( Alwan and Roberts, 1995). On one hand this invalidates the statistical process control (SPC) chart.

On the other hand, serial correlation in the process variable enables the fitting of a time series model, from which predictions can be made. Feedforward control requires that out of control information, based on systematic process variations be available ahead of time. This type of information will be contained in the systematic serially correlated fitted values from a time series model. However, it is very important that the fitted values not be biased. Therefore, care must be taken to correctly specify the time series model. Otherwise, the residuals from the time series model may also be serially correlated, resulting in no real progress. Alwan and Roberts (1988) and Alwan and Radson (1995) suggested the use of a time domain time series model to isolate IID residuals from non-IID fitted values. However, Alwan and Radson (1995) reported difficulties when the process variable contained cycles. They were unable to automate the procedure to reliably fit a time domain model to the data. In this paper, we use the moving window spectral (MWS) time series model. It is a frequency domain model capable of automatic representation of trend and several hidden cycles. Details of the MWS method may be found in Ridley (1994a,1998,1993,1994b).

In this paper, we discuss the application of the MWS model and a neural network (NN) model to process control, and the possibility which it creates for reducing the need for feedback trim, reduction in process variability, and quality beyond that currently conceptualized by TQM. A NN is trained to recognize unnatural patterns that might occur in the process variable (see Ramirez-Beltran and Llaugel, 1994). Ultra high signal to noise ratio fitted values from the MWS time series model are then fed to the NN. The NN detects the formation of unnatural patterns. This forms the basis for corrective action to be taken. In this way the process never actually goes out of control.

AN MWS-NN MODEL BASED SPC SYSTEM

An overview of the integrated control system is given in Figure 1. There are two reasons why the process may become out of control . Either the inputs to the process are different from what is expected or the conversion process itself develops a component malfunction. Disturbances to the primary inputs are either measurable or unmeasurable (see figure 1 - bottom left section). An example of a disturbance is a bad batch of raw materials. Unmeasurable disturbances will enter the process undetected, resulting in defective output from the process. Random but measurable disturbances are fed forward to a controller immediately as they occur. The controller then manipulates the input so as to compensate for the disturbance. Any remaining deviations from set point are corrected by feedback trim. Dynamic compensation via feedforward control and feedback trim maintain the system in control. Feedforward control reduces the amount of feedback trim that is required. For further discussion on the application of feedforward control see Fisher Controls(1998).

Figure 1. Moving Window Spectral Model Based Feedforward Process Control.

oiwnhdfoiDS[N

Malfunctions which develop within the conversion process will only appear in the process output (see Figure 1 - top right section). In order to detect these a short history of output measurements are analyzed by the MWS time series model. The fitted values from the MWS model are plotted on a common cause chart, while residuals are plotted on a special cause chart. These charts provide an opportunity for visual inspection of the process.

The fitted values and the residuals are also fed to a NN for analysis (see Figure 1 - top left section). The NN network is set to detect any breaches in the special cause SPC chart, and to alert the system operator accordingly, for corrective action. Special causes are one of kind. No systematic process correction is possible. For example, in the case of a bad batch of raw materials, the supplier should be notified. If it occurs more frequently than is acceptable, a different supplier must be used. The fitted values are systematic and therefore predictable. Predictions by the NN are fed to a feedforward controller. The controller then manipulates the input so as to commence gradual compensation of this systematic disturbance, ahead of time. The objective is to prevent the system from ever becoming out of control. In a fully automated system, human intervention is minimized. However, continuous monitoring by an operator is possible via the dual system of special cause and common cause SPC charts.

CONCLUDING REMARKS

A seamless integrated moving window spectral (MWS) time series and artificial intelligence neural network (NN) feedforward model based statistical process control system (MWSNN) was described. This integrated procedure is performed by a the computer program FOURCAST (Ridley,1998), with a call to the NN. FOURCAST automatically generates statistical process control charts for residuals and for fitted values, marked with 1, 2 and 3 sigma limits, providing visual inspection capability and human backup to the NN for difficult cases where automation may fail.

Further research may incorporate an assignable cause interpreter expert system (ES) into the artificial intelligence system, with the diagnostic capability to identify and locate specific components of the production system, which may be failing. The component list could include people, equipment and/or processes. The interpreter would have to be production plant specific. The output from the NN would be fed to an ES which is trained on the particular plant.

REFERENCES

Alwan, L.C., and D. Radson. “Implementation Issues of Time-Series Based Statistical Process Control”. Production and Operations Management. Vol.4, No. 3 (1995). pp. 263-276.

______and H. V. Roberts. “Time-series modeling for statistical process control”. Journal of Business and Economic Statistics. Vol.6, No. 1 (1988). pp. 83-95.

______and H.V. Roberts. “The problem of misplaced control limits (with discussions)”. Applied Statistics. Vol.44, No.3 (1995). pp.269-278.

Fisher Controls. Feedforward Control (student guide). Fisher Controls Educational Services, Marshalltown, Iowa, 1988.

Ramirez -Beltran, N.D., and F. Llaugel. “The ratio control chart and pattern recognition”. ASQC 48th Annual Quality Control Proceedings. (1994). pp.583-590.

Ridley, A.D. “Variance stabilization:A direct yet robust method”.Rev.of Bus.Vol.15,(1993).pp.28-30.

______“The global univariate moving window spectral method”. Report number FSU- SCRI-94-17. Supercomputer Computations Research Institute, Florida State University, Tallahassee, FL 32306, USA. (1994a). Email: . Tel(850)644-1010.

______“A model-free power transformation to homoscedasticity”. International Journal of Production Economics. Vol.36, (1994b). pp. 191-202.

______FOURCAST-Multivariate spectral time series analysis and forecasting(mws). EMC, Box 12518, Tallahassee, Fl 32317-2518, USA. 1998. URL= http://www.fourcast.net/

Western Electric Co. Statistical Quality Control Handbook. Western Electric, Ind., IN. 1958.

Proceedings of the Tenth Annual Conference of the Production and Operations

Management Society, POM-99, March 20-23, 1999 Charleston, S.C.