PRBS sequence

For tracer experiments periodic sequences of pseudo random binary signals (PRBS) can be used, which show good results after a shorter measuring period (Havlicek and Cermak, 1977 Unbehauen and Funk, 1974). A PRBS sequence consists of a stochastic series of rectangular tracer pulses (concentration c, time interval At). The sequences recur with a period T. Each period is divided into N time intervaUs T = N-At). 

As stated in the previous section, the major reactant feed was chosen as the manipulated variable. In the trial this feed was subjected to a pseudo-random binary sequence (PRBS) signal in an open loop operation of the process. The results of the trial, plotted in Fig. 2, show a strong -- but delayed -- cross-correlation between the manipulated feed rate and the reactor temperature. Using techniques described by Box and Jenkins (2), a transfer function relating the manipulated variable to the control variable of interest can be developed. The general form of this transfer function is... 

An alternative method (developed in the Eli Lilly research laboratories), entails inserting a nucleotide sequence coding for human proinsulin into recombinant E. coli. This is followed by purification of the expressed proinsulin and subsequent proteolytic excision of the C peptide in vitro. This approach has become more popular, largely due to the requirement for a single fermentation and subsequent purification scheme. Such preparations have been termed human insulin prb ... 

A very popular sequence of inputs is the pseudorandom binary sequence (PRBS). It is easy to generate and has some attractive statistical properties. See System Identification For Self-Adaptive Control, W. D. T. Davies, London, Wiley-Iflterscience, 1970. 

Figure 7 Addition of noise to the detector signal of a single pulse chromatographic run and a run using a PRBS injection sequence.

A special kind of random noise, pseudo random noise, has the special property of not being really random. After a certain time interval, a sequence, the same pattern is repeated. The most suitable random input function used in CC is the Pseudo Random Binary Sequence (PRBS). The PRBS is a logical function, that has the combined properties of a true binary random signal and those of a reproducible deterministic signal. The PRBS generator is controlled by an internal clock a PRBS is considered with a sequence length N and a clock period t. It is very important to note that the estimation of the ACF, if computed over an integral number of sequences, is exactly equal to the ACF determined over an infinite time. 

The pRb protein is a nuclear phosphoprotein of ca. 100 kDa. The domain structure of pRb is shown in Fig. 13.13 (review Riley et al., 1994 Weinberg, 1995). The pRb proteins possesses numerous Ser/Thr phosphorylation sites, binding sites for the transcription factor E2F, for the viral oncoproteins TAg, ElA and E7 (see Chapter 14) and a non-specific DNA binding site. Furthermore, an N-terminal sequence section has been identified in pRb protein that is required for oligomerization. 

Fig. 13.13. Domain structure of the retinoblastoma protein pRb. The phosphorylation sites (P) of pRb and the localization of the sequence sections necessary for interaction with viral oncoproteins and with the transcription factor E2F are shown. In addition, an oligomerization domain and a DNA binding domain can be identified.

A segment of the A genome containing the Prm and nearby PL, PR, and PRb. promoters. Transcripts that can originate from the three promoters are shown by the wavy lines. The PRe and -PRM require different activators for expression. The transcript from PRF is 10 times more effective in cl expression because it has a Shine-Dalgarno sequence for ribosome attachment. 

Substrate binding involves interactions with both the CDK and the cyclin. Whereas the CDK contacts mainly the residues surrounding the target Ser/Thr of the substrate, interactions between distinct sequence elements of the cyclin and the substrate have been shown to contribute to substrate specificity of cyclin-CDK complexes as well. The region of the cyclin responsible for this interaction is also involved in the binding of CDK inhibitors and other regulatory proteins such as the retinoblastoma protein, pRb (see Section 13.4.2). 

Fig. 5 C-terminal Cdk-binding domain of p21 conjugated with antennapedia PTD suppresses pRb phosphorylation and induces Gj/S cell cycle arrest. (A) Peptide sequences of fusion polypeptides composed of C-terminal Cdk-binding domain of p21 (p21141.160 [peptide I] or p21i54.i6o [peptide II]) and antennapedia PTD boxed). (B) pRb became hyperphosphorylated between 12 and 15 h after serum was added to starved HaCat cells, but in the presence of peptide I or II remained hypophosphorylated. pRb = hypophosphorylated, pRb = hyperphosphorylated. (C) Cell cycle distribution of HaCat cells after culture in D-MEM medium containing 10% PCS alone or 10% PCS with peptide I or II. Note Peptide I and II inhibited the phosphorylation of pRb and induced Gi/S cell cycle arrest. This figure was modified from Ref [120] therefore, see Ref. [120] for details...

Stochastic identification techniques, in principle, provide a more reliable method of determining the process transfer function. Most workers have used the Box and Jenkins [59] time-series analysis techniques to develop dynamic models. An introduction to these methods is given by Davies [60]. In stochastic identification, a low amplitude sequence (usually a pseudorandom binary sequence, PRBS) is used to perturb the setting of the manipulated variable. The sequence generally has an implementation period smaller than the process response time. By evaiuating the auto- and cross-correlations of the input series and the corresponding output data, a quantitative model can be constructed. The parameters of the model can be determined by using a least squares analysis on the input and output sequences. Because this identification technique can handle many more parameters than simple first-order plus dead-time models, the process and its related noise can be modeled more accurately. 

In order to obtain adequate data estimation for the reaction rates, PRBS input sequences in Fb and Tr were introduced, using a sample time of 1000 sec. Operation conditions and outlet concentrations were register to train the neural networks. The model update that consists of the NN adaptation was carried out using a second order recursive algorithm. The best NN structures were found by a systematic training procedure, considering outlet concentrations of A and B components and the reactor temperature as the inputs to the networks. Finally, networks with one hidden layer with four nodes and sigmoidal activation functions were selected. 

Figure 4 The shear transition from a 4 to a 3 net (above), the crystal structures of PrB-II and -III (middle), as weU as of (bottom) Prl2-IV and -V in equivalent projections showing the layer sequences...

PRBS (pseudo random binary sequence) signal is widely used in identifying the unknown parameters of a system because of its good statistic properties, and the possibility of reducing the noise to the least level. 

---