DE10034645A1 - Monitoring and control process for biological sewage plant, useful for treating sewage of varying composition from industrial, agricultural, biological or communal source, uses neuronal network computer to generate control signals - Google Patents

Abstract

In a process for monitoring and influencing operation of biological sewage plants and stages for treating sewage (mixtures) by determining its properties, especially suitability for treatment, before it is passed into the treatment works, the sewage is mixed with activated sludge, nutrient substrate and oxygen and various parameters are measured and used for training a neuronal network computer to generate output values for control. Process for monitoring and influencing the process in biological sewage plants and stages for treating sewage and mixtures of varying composition involves determining the properties of the sewage, especially its suitability for treatment, before it is passed into the treatment works. The sewage in a monitoring vessel is mixed with activated sludge and nutrient substrate and oxygen is added. One parameter, the oxygen consumption of the microorganisms, which are present in the activated sludge and consume the nutrient, is measured in relation to other initiated parameters and evaluated electronically for 5-15 minutes, until the degree of oxygen saturation ensuring optimum consumption by the microorganisms is reached. The parameters, chemical, biological and physical oxygen partial pressure values, temperature, electrical conductivity of the amount of dissolved organic carbon, turbidity, pH, mixing ratio of sewage of various origins and time of day or year, are measured periodically in parallel. The measured values are converted to time-dependent data and transferred to a computer with a neuronal network and sets of data for recognizing and evaluating the relevant and irrelevant measured values for input into the plant, are trained over a long period. The relevant and irrelevant data are processed directly to output values. The suitability is determined by the extent of negative deviations of the gradient of the oxygen consumption curve prognosticated by the neuronal network and used for regulating the monitor before sewage treatment. An Independent claim is also included for the equipment used in this process.

Description

The invention relates to a method for monitoring and controlling the process sequences for determining the properties of wastewater of different types setting from industrial, agricultural or municipal areas as well as a Device for performing the method.

It is known that wastewater from industrial, agricultural or municipal Areas in their composition are subject to large fluctuations. So can for example discontinuous production processes, changes in manufacturing process or accidents in chemical plants at a short or long term Change the quality of the waste water. Heavy rains would also Waste water composition, for example in municipal wastewater treatment plants rivers. Discharges of wastewater that are subject to such fluctuations to overloads or poisoning of the downstream sewage treatment plants. The consequence is a death or flushing away of the biomass, which degrades its content and Serves pollutants in the wastewater, from the biological clarification stage of the wastewater plant, which reduces the dismantling performance of the system and the risk of exceeding the permissible pollutant limit values for the discharge of the treated wastewater into the Receiving water exists.

JP 100 43 788 A discloses a control method for treatment facilities of polluted wastewater. The disclosed control device for a biological Sewage treatment plant works on the principle of determining the phosphate or Nitrogen concentration with temperature, the data of the volume flow in one neural network can be fed, which then as the direct control the Aeration intervals of the return sludge ratio and the excess sludge removal regulates. The control influences the process control of the plant with the The aim is to adhere to the values of the proportion of phosphate and the amount of nitrogen. The font does not disclose solutions in the field of toxicity detection or its use or the use of neural networks for this limited area. The Control is used within the plant and can detect toxicity do not signal waste water flowing into the system.  

Joachim Hansen, Jörg Ramacher, "Fuzzy Control and Neural Networks in the Wastewater treatment "Correspondence sewage 46/1999, No. 2 p. 168; 169, provides the Application of neural networks based on fuzzy control for controlling the Operation of existing biological sewage treatment plants. It deals with the Optimization of nitrogen and phosphorus elimination. When performing the Various wastewater parameters are determined in the process, and the parameters obtained are determined Information evaluated and using known automation techniques to control the Processes in the sewage treatment plant. The volume flows that Residence times and the aeration rates of the wastewater to be treated during the Process flow measured and weighted. The use of neural networks has the aim is to use conventional control technology to reduce the Nitrogen and phosphate concentrations slipping in during the wastewater treatment regulate. The document gives no information, such as recognizing the toxicity in the inflow the sewage treatment plant should be carried out and how it can be possible the toxicity to be recognized smoothly and timely measures to maintain the operational safety of the Initiate plant for wastewater treatment.

M. Köhne puts in "Optimization of wastewater treatment processes in sewage plants on the Basis of continuous measurements "; Wasser und Boden 48. JG 9/1996, pp. 8-12, show how a continuous measurement method for controlling wastewater treatment processes in the associated systems can run. Above all, it shows how the Evaluation of the measurement data essentially to reduce the legally relevant Waste water ingredients in the regime of strictly reduced energy consumption to be carried out. For example, the article states that a biological Nitrogen elimination is done by on-line measurement of nitrogen and oxygen maximum nitrogen degradation with minimal oxygen consumption can be achieved. Optionally, the content of the publication affects implementation of control processes including neural networks. How toxicity in Inlet of the sewage treatment plant recognized for the process regime and to influence the Treatment process is not used here and neither Subject of the continuous measurements of the Wastewater parameters.  

The same author poses in "On-line measurement methods in wastewater engineering"; Correspondence sewage 10/1993 40th JG, pp. 1628-1638, a process of on-line Measurement with the following principles.

Monitoring and control of the function of the sewage treatment plant and its cleaning performance with the processes: measuring, registering, documenting.

Manual and automatic influencing of the wastewater treatment process with the Objective: controlling, regulating and optimizing the system function.

Obtaining general information from the overview for the detection of continuous and variable functioning of the sewage treatment plant in the sense of analyzing, identifying, Model and simulate.

The publication continues through a comparison in the context of practical Findings with advantages and disadvantages more continuously and discontinuously Measurement methods apart. TS measurements, BOD measurements and Ammonium measurements that are seen individually. A Detection of toxicity is in no way the subject of the reference Process Control Procedures. Every concrete one is missing in this publication Regarding the use of artificial intelligence, i.e. of neural networks, especially with a toxicity or property detection on the wastewater treatment plant incoming wastewater.

When operating biological sewage treatment plants, there is an imperative that protect existing biomass from toxic discharges. To this end waste water from the various areas of origin pass through before entering the Sewage treatment plant often called "toxicity monitors". This is Measuring devices that work continuously or discontinuously and statements about the safety of the wastewater to be discharged for the biological activity of the Sewage treatment plant. If toxic influences occur, e.g. B. inadvertent Introduction of toxic ingredients, a warning message is issued so that the Waste water can be diverted into an emergency basin. All currently The available toxicity monitors are based on the measurement of oxygen consumption. This The detection process quantifies the metabolic performance of the microorganisms the rate of decrease in oxygen concentration in one Reaction tank in which wastewater and certain microorganisms are mixed.  

The result of this measurement is an oxygen consumption curve, which is shown below by way of example.

With a constant biomass concentration in the reaction vessel, the rate of oxygen removal - that is, the steepness of the depletion curve dO ₂ / dt - represents a measure of the metabolic activity of the microorganisms. However, the problem with this approach is the fact that a relationship between oxygen consumption and metabolic activity is only possible under reproducible conditions is valid. In addition to the biomass concentration, z. B. still the pH, the salinity (conductivity), the temperature and the substrate concentration on the oxygen consumption. Another important influencing factor is the basic vitality of the biomass used under normal conditions. T. is subject to considerable fluctuations.

Because the listed chemical-physical parameters of wastewater in practice cannot be influenced, at least when evaluating the Find oxygen consumption. Without paying attention to the individual Wastewater parameter is a satisfactory statement about the suspected toxic  Effects of waste water not possible. The wastewater quality is with the current device types, as shown later, but not on the market considered. The flawless functioning of such toxicity monitors is therefore essential Wastewater quality largely constant over long periods. So it is from a condition that occurs very rarely in practice.

At present, changing wastewater compositions are being identified Single parameter measurements carried out, electronically recorded and individually extrapolated. Such measurements are often very lengthy and right to evaluate confusing. The more parameters included in the assessment of wastewater quality the more inaccurate the total becomes in the conventional data comparison result, since constantly changing wastewater compositions make it necessary to adapt individual Complicating parameters to the overall database. Add to that individual Waste water constituents are often difficult to analyze and often only that Interaction of several components to impair cleaning lead process. Furthermore, there is a risk that between the investigation of the Waste water and the issuance of a warning takes too long a time before using appropriate means to respond to a possible disturbance.

Biological monitoring systems are also known which measure the breathability of microorganisms in biological media via individual parameters.
Kümmerer, Friedrich
"Operation monitoring of the sewage treatment plant using on-line BOD measurements"
Correspondence sewage No. 7/99, pp. 1086-1093

The method shown is based exclusively on the oxygen consumption measurement. The results are not evaluated on the basis of neural networks. Furthermore, the proportions of activated sludge and waste water used here are disadvantageous for the evaluation.
Guhl; Gode; Peter
"Malfunctions of biological sewage treatment plants caused by chemicals"
Comparison of the limit concentration with results in the oxygen depletion test; From the water, 72 165-173

The method shown is based exclusively on the oxygen consumption measurement. The results are not evaluated on the basis of neural networks.  

Basically, this work only compares the limits of laboratory wastewater treatment plants in the presence of substances suspected of being toxic with a simpler oxygen depletion test. A forecast of the toxicity with the issue of a warm message was not the aim of this investigation. The transferability of the results obtained to a real wastewater treatment plant is also questioned.
Pilz, Ulrich
"Method and device for the determination of the nitrificants - toxicity"
DE 196 40 334 A,

The procedure described in the paper is based solely on investigations toxic effects on nitrifying bacteria. This group of organisms only represents represent a small part of the microorganisms involved in COD degradation, but there are none Overview of the entire process during the cleaning process. An evaluation the basics of larger overviews of measured values are not recognizable.

To determine a possible toxicity, the oxygen consumption of Microorganisms in contact with the wastewater to be investigated. The Toxicity is the measure of the decrease in oxygen consumption, or is the amount Fresh water specified, which must be supplied to the measuring system in order to Remove breathing inhibition. In the presence of toxins, breathability can decrease be significantly reduced. However, it depends not only on inhibitors, but also of the environmental conditions, such as temperature, salinity, pH, concentration in nutrients, concentration of microorganisms and solids content, so that often a wrong conclusion is drawn from poor breathability. A low substrate content, for example, causes only a low oxygen consumption, d. that is, the measured change in breathability is very small. But that's from it not recognizable to what extent toxic substances have affected the entire system. The following toxicity monitors are known to determine the toxicity of waste water:

Type designation

TOXalarm pro, LAR Analytik und Umweltmesstechnik Berlin
STIPTox adapt, Stip GmbH, Groß Ulmstadt
RODtox, UPM Umwelt Pollution Messtechnik GmbH, Langgöns

From this overview it is clear that all products shown here are from one constant oxygen consumption, below which a Toxicity warning is issued. The real wastewater properties are in not considered in any way. This leads to the practical operation of such systems frequent triggering of false alarms.

The invention has for its object a method and an apparatus for Monitoring and control of process flows for the determination of the properties of wastewater from industrial, agricultural or municipal areas create a continuous, flexible and short-term quality assessment the waste water is reliably possible and the occurrence of faults in the Process flow of the system can be avoided extremely quickly with countermeasures.

According to the invention the object is achieved in that with a method for Monitoring and influencing processes in biological sewage treatment plants and Clarification stages for the treatment of wastewater and wastewater mixtures are changing Composition a sample of the wastewater to be monitored in one Monitoring tank is mixed with activated sludge and growth substrate, to this Mixture supplied oxygen and then in the mixture as a parameter  Oxygen consumption of microorganisms depending on other parameters is determined and evaluated with an artificial neural network.

The invention is advantageously embodied in that from the time of reaching a Degree of oxygen saturation, at which an optimal oxygen for the microorganisms consumption is guaranteed to vary over a period of 5 to 15 minutes Parameters are determined in parallel, the measured values from measurement data converters in converted time-dependent records and the records via electronic Interfaces to a computer with a neural network are transferred. The records from there, taking into account the recognition of their properties, for the Operation of the facility to be suitable or not suitable, quickly and directly Output values processed. The determined output values of the neural network are for the control of the monitoring device of the devices for waste water cleaning used.

An essential content of the invention is that the neural network determined the Data records are considered as input data according to their weighting and mathematically with a threshold that is a threshold for accepting the Represents the presence of toxicity or non-toxicity of the wastewater, whereby from this comparison output values are generated with which the Monitoring device or devices in front of the sewage treatment plants and clarification stages directly can be controlled and a statement about the quality of what is to be introduced into the system Sewage is given. This way, complex systems with qualitative different parameters, without knowledge of functional relationships, between the parameters.

Advantageously, by determining the parameters in the monitoring wastewater obtained data sets mathematically weighted to whether they are of a suitable size. Likewise, measuring technology falsified measured values or measured values from the determination of parameters that are in the Waste water mixture influence each other in such a way that adulteration occurs weighting. In the neural network the ability to weight the Data records before the actual start of the monitoring measurements with measured values generated from wastewater during normal, trouble-free operation  were. If individual data do not exist the weighting, they are from the neural network filtered out and no longer considered in the further course of the calculations. According to the invention, the threshold value of the neural network is created using data records, which were obtained separately from waste water. The threshold thus represents Monitoring method is a value, when falling below or exceeding the Normality of the waste water is generated as an output value.

It is essential for the invention if different chemical, biological or physical measured values can be determined in parallel at the same time preferably around the oxygen partial pressure, the temperature, the electrical Conductivity, the amount of dissolved organic carbon, the turbidity, the pH Value, the mixing ratio of the wastewater quotas to one another for wastewater of different origins or around the time of day or season, in the case of waste water quotas are introduced into a sewage treatment plant or clarification stage. The parameters are with low technical effort easily and quickly determined. The simple and quick Determinability of the parameters is an important prerequisite for the shortest possible Measuring cycles. Short measuring cycles are in turn for safe operation and Monitoring in front of a wastewater treatment plant is necessary so that, if necessary, toxic measured values enough time for adequate control of the sewage treatment plant remains. The dead time between the waste water analysis and the discharge of a Warning whether the wastewater should be sent to a sewage treatment plant exceeds one Period of 15 minutes is not, but can already be output in seconds. An advantageous embodiment of the invention is that activated sludge is used which is also for the cleaning stage of the respective biological sewage treatment plant or Clarification stage is used. Special activated sludge for the monitoring process are not required. There are a number of differences in the activated sludge most microorganisms for the purification of the waste water must be suitable. The used in biological treatment plants and treatment stages Activated sludges ensure a reliable course of the parameter determination.

The microorganisms need oxygen to be in the wastewater as a prerequisite for life Break down ingredients. When the ingredients are broken down, the microconsume organisms more oxygen than without degradation, so that there is a significant decrease in the Oxygen concentration in the wastewater comes. The decrease in oxygen concentration  depending on the time is a measure of the degradation activity of the microorganisms. Is disrupted the degradation activity of the microorganisms, d. that is, only very small or no reductions in the oxygen concentration measured in the wastewater, so various causes may be responsible. The oxygen depletion activity of the Microorganisms can be caused by toxic substances, by suboptimal living conditions, such as temperature, pH deviations u. a. or by the presence of one too low concentration of substances that can be broken down by microorganisms, be inhibited.

Immediately before the start of parameter determination, the mixture of wastewater, Activated sludge and growth substrate according to the invention oxygen up to one Degree of saturation of 50-95% fed. It is of particular advantage solution according to the invention if the mixture has an oxygen saturation degree of 80 at the beginning of the measurement, so that the microorganisms on Have an optimal oxygen depletion potential at the beginning of the measurement. On Optimal consumption potential ensures good living conditions for the micro organisms. This also means that the parameter oxygen is constantly tracked consumption over a wide measuring range to generate large data sets possible before the oxygen concentration in the mixture by consuming the microorganisms it becomes small that the oxygen concentration can no longer be measured. The Oxygen can be added to the mixture as a pure gas or together with inert gases become.

Another embodiment of the invention is that the mixture of waste water and Activated sludge doses a growth substrate that does not damage the microorganisms that the microorganisms can continuously break down. The lack of content and pollutant levels in the wastewater result in a lower degradation activity of the Microorganisms, however, gives no information about the cleaning properties of the Sewage. Only the addition of the food substrate ensures that the Microorganisms even if the wastewater with insufficient content and Proportion of pollutants that actually determines the degradation activity of the microorganisms, is burdened.

It is advantageous if nutrients are used as the growth substrate Microorganisms are adapted, preferably a protein or protein mixture,  a lipid or lipid mixture or a carbohydrate or carbohydrate mixture is used becomes. It is an advantage when using the solution according to the invention that the Proteins, lipids or carbohydrates are available at any time through purchase and for a high cultivation activity of the microorganisms are well suited. One at any time Reproducible and stable consumption activity is essential for maintaining well-structured Records an important requirement.

It is an advantage of the method according to the invention that at the beginning of the actual Oxygen consumption measurement the relevant physico-chemical parameters of the Waste water can be determined in the reaction vessel. In addition, more wastewater Characteristics that are available to the respective operator and for character terization (origin, volume shares, etc.) are included in the data record accepted. Furthermore, a so-called background substrate is added to the wastewater / Biomass mixture dosed to ensure that even with low substrate concentrations (e.g. strong Dilution with rinsing water or similar) the measurable decrease in the oxygen concentration to secure. Otherwise, as with products from other manufacturers, this would happen several times happen - there is a toxicity warning due to lack of measurable oxygen consumption. In a training phase lasting several weeks, there will be several during normal operation of the sewage treatment plant thousand records, d. H. Combinations of measurement parameters and the associated one Oxygen consumption, determined. This creates relationships in a neural network "learned" between the properties of wastewater and oxygen depletion. To Completion of the training phase can then be expected with high accuracy Oxygen consumption of every wastewater - even with different partial flows composition - to be predicted. With this default value stands for everyone Wastewater or wastewater mixture an individual parameter available with their With the help of the real measured value, it can be directly compared and precisely evaluated. in the later routine operation, the constantly new measurement data will become Training of the neural network used. This creates a self-learning system which is able to adapt to changes in the wastewater composition in the course of Adjust time and the suitability of the wastewater for discharge into the plant as Always redefine deviation from the "normal" setpoint.

With the method, an effective recognition of the suitability for introduction is advantageous get into the plant by safely determining the "normality" of the wastewater  can and the unsuitability due to the extent of the conspicuous (negative) deviation from the predicted normal value, which suggests toxicity as a hazard becomes.

The device according to the invention for carrying out the monitoring method and control of processes before biological treatment plants and treatment stages for the Treatment of waste water of different compositions consists of one Monitoring tank to hold the waste water, the activated sludge, the Depletion substrates and oxygen. For dosing the wastewater, the Activated sludge, the drainage substrate and the oxygen are before the Monitoring container dosing devices arranged. The monitoring tank has a mixing device that after the addition of all components in the Containers mixed together. There is also a cleaning procedure on the container attached, through which the used mixture of the Monitoring container is removed.

The invention receives a meaningful training in that the monitoring container at least two different measuring probes are arranged, with which in the wastewater the different chemical, biological or physical parameters several times can be measured. There are at least two for the correlation in the neural network Parameters required that have an effect in the mixture system to be examined exercise. The more parameters are measured, the more data sets can be used for the neural network are made available, reflecting the influence of incorrect data reduced to the entire monitoring process.

According to the invention, the measuring probes are connected to measuring data converters in which the values determined during the parameter measurement are converted into electronic data records so that the qualitatively different, but time-dependent Data records can be processed in the neural network. The measurement data converters are for transferring the data records via electronic interfaces with the computer connected. The computer is equipped with a neural network that consists of the Records output values generated by the control of the monitoring device and serve to assess wastewater quality. The calculator is advantageously used for Control and repeated repetition of the monitoring processes with the Measuring probes, the mixing device and the metering devices electronically connected.  

In a further advantageous embodiment of the invention, the computer in which the datasets are generated with the aid of the neural network, which the Control of wastewater treatment processes can be used with the devices Wastewater treatment plant to control the cleaning processes electronically connected.

The invention will be explained below using an exemplary embodiment. In the associated drawing show:

Fig. 1: The diagram of the measurement value acquisition and processing

Fig. 2: Representation of the measurement and evaluation process

Fig. 3 The location and evaluation of measuring points in the neural network

The method is to be explained on the basis of the schematic representation of the device according to FIG. 1. Waste water 1 , activated sludge 2 , growth substrate 3 and oxygen 4 are in suitable storage containers before the start of the first cycle of the monitoring process. With the aid of the metering devices 12 , the components are placed in the order 1 , 2 and 3 in succession in the monitoring container 6 and mixed with one another with the aid of the mixing device 7 . With further mixing, oxygen 4 is supplied in the form of air until an oxygen saturation level of 80% is reached in the mixture. After the oxygen supply has ended, the measuring cycle begins. With the aid of the measuring probes 5, the measured variables oxygen partial pressure, temperature, electrical conductivity and pH value are determined in parallel at intervals of 15 seconds over a period of 10 minutes. After the measurement value recording has ended, the monitoring tank 6 is emptied via the cleaning process 8 and rinsed with new wastewater 1 . The rinse water is also removed from the monitoring tank 6 and a new measuring cycle can begin. The measured values determined are forwarded to the measurement data converter 9 and converted into time-dependent data records. The data records are transmitted via the electronic interface 10 to the computer 11 , which has an artificial neural network 18 in which they are first weighted, then mathematically related to one another and compared with a threshold value. The threshold value and the ability for weighting were generated for the neural network before the start of the monitoring process with data records that were obtained separately from the assessed wastewater. By comparing the processed data sets with the generated threshold, the neural network issues a yes / no decision on the suitability of the waste water for discharge into the system as a signal. The output signal triggers the start of a new measuring cycle in a mathematical program with control 14 of the measuring probes 5 , the mixing device 7 and the dosing devices, so that several measuring cycles can be carried out in succession.

The course of the method according to the invention shows the representation of the measurement and evaluation method according to FIG. 2. The degree of saturation of the waste water with oxygen is shown in a diagram 15 . After reaching the highest degree of saturation, the consumption measurement starts at approx. 4 minutes. The pH value, the conductivity, the temperature and / or other required characteristic values are recorded and sent to the computer with a neural network according to direction arrow 16 '. After a training phase, the computer 11 calculates the forecast of expected oxygen consumption as a function of the wastewater parameters. At the same time in FIG. 5, the slope of the oxygen consumption curve is recorded in the course of the consumption measurement in the range between 5 and 14 minutes and supplied to a comparison unit at 17 . In comparison, the forecast is compared with the measurement. If there is an overlap, i.e. the identity of the calculated values of the slope of the curve of oxygen consumption with the measured values, there is a supply line at 17 "to the wastewater treatment plant and if there is a mismatch with the expected oxygen consumption depending on the wastewater parameters, there is a warning and one at 17 " Waste water is diverted from the normal flow direction.

Fig. 3 illustrates in a three dimensional coordinate system selected on the axes of three measurement values of the measurement time point h, the temperature T and the pH are plotted against. In the neural network 18 fitted therein, which is symbolically represented as a matrix, there are measurement points 19 ; 20 classified. The measuring points 19 are recorded continuously during the weighting of the measured values in the measurement run of the method according to the invention, processed in the computer 11 and transferred to the neural network 18 . The neural network 18 signals that the measuring points 19 are normal and thus the assessed wastewater in the overall assessment of its composition is suitable for introduction into a wastewater system. The measuring point 20 is striking in its position and analysis meaning. It 20 stands out from the band of the matrix of measurement points 19 recorded by the neural 18 as "normal" and has so far been assessed as normal and unobtrusive. The evaluation result "abnormal" appears as a signal and the wastewater is not discharged into the wastewater treatment plant until further signals appear. Only when all of the measuring points are within the normal range of the total values weighted from the matrix when the wastewater is assessed is the wastewater reintroduced into the wastewater system. For better clarity, three measured values h, T and pH have been assumed in the exemplary embodiment. Of course, according to the invention it is possible to use more than three measured values and to implement these as measured points and compare them with the empirical values of the neural network 18 . Since the neural network 18 is self-learning, the value of the composition of the wastewater referred to as "normal" can shift during the continuous evaluation of the measured values over time, but in no case reaches the limit of the general non-processability when discharging into the wastewater system. Since the measured values are compared with short time pulses, the continuity of the determination of the suitability for an introduction of the inflowing wastewater in front of the wastewater system is very great.

List of the reference symbols used

1

sewage

2nd

Activated sludge

3rd

Food substrate

4

oxygen

5

Measuring probes

6

Monitoring container

7

Mixing device

8th

Cleaning process

9

Measurement data converter

10th

electronic interface

11

Computer with a neural network

12th

Dosing devices

13

Control of the wastewater treatment plant

14

Control of the monitoring device

15

Diagram of the process of the O2 consumption

16

;

16

';

16

"Directional arrows

17th

;

17th

'Arrows

17th

";

17th

'''Directional arrows

18th

neural network

19th

Measuring point normal

20th

Measuring point striking
H; T; pH reading

Claims (15)

1. Procedures for monitoring and influencing biological processes Wastewater treatment plants and stages for the treatment of wastewater and Waste water mixtures changing composition by determination the properties of the waste water, in particular its suitability for Further processing before the introduction into the processing plant, whereby the Wastewater, which is in a monitoring tank, activated sludge and the growth substrate are mixed, oxygen is added to this mixture is then a parameter in it as the oxygen consumption of microorganisms that are in the activated sludge and consume the growth substrate, in Dependency on other initiated parameters determined and electronically is evaluated and when the oxygen saturation level is reached, in which optimal oxygen consumption is guaranteed for the microorganisms, over a period of five to fifteen minutes as a chemical parameter, biological and physical measurements of the oxygen partial pressure of the Temperature, the electrical conductivity of the amount of dissolved organic Carbon, the turbidity, the pH value, the mixing ratio of the Wastewater quotas for each other for wastewater of different origins, daily or season, a sewage treatment plant or clarification stage, recorded in parallel the measured values from a measurement data converter into time-dependent Records converted and the records via electronic interfaces too transferred to a computer with a neural network, the data records for detection and Evaluation of for a determination of the discharge into the sewage system relevant and irrelevant measured values, trained over longer periods, relevant or irrelevant measurement values processed directly into expenditure values, the determination of suitability by the extent of negative deviations of the slope of the curve of the curve predicted by the neural networks Oxygen consumption is detected and used to control the monitoring device in front of the Wastewater treatment is used.   2. The method according to claim 1, characterized in that before the start of Parameter determination oxygen up to a degree of saturation of 50-95% of the Waste water mixture is supplied. 3. The method according to claim 2, characterized in that before the start of the Parameter determination oxygen up to a degree of saturation of 80% of the Waste water mixture is supplied. 4. The method according to claims 1 to 3, characterized in that Activated sludge is also used for the cleaning level of the respective biological sewage treatment plant or clarification stage is used. 5. The method according to claims 1 and one or more of the following Claims, characterized in that the wastewater mixture in connection a background substrate is metered in with the biomass. 6. The method according to claims 1 to 5, characterized in that for Detection of ingredients in the wastewater parallel the neural network data records determined with one another as input data according to their weighting taken into account and mathematically with the threshold value, which is a limit of Normality of oxygen consumption is in a comparison, taking from this Comparative output values are generated that give a statement about the suitability of the Waste water to be discharged directly into the sewage treatment plant and treatment stages. 7. The method according to claims 1 to 6, characterized in that the Threshold value of the neural network is formed from data sets with which one Information about the curve of oxygen consumption at Normal conditions is conveyed by means of those in the monitoring process a value is output to the sewage treatment plant if the value is fallen short of or exceeded in the real measurement of the wastewater a value for the unsuitability of a Discharge into the sewage system is obtained. 8. The method according to claims 1 to 5, characterized in that for the Ensuring a good growth activity of the microorganisms that a  Prerequisite for the creation of usable data records for the neural network is nutrient adapted to microorganisms as a growth substrate in the form of a background substrate. 9. The method according to claim 8, characterized in that the cultivation substrate Background substrate is a measurable at low substrate concentration Secure decrease in oxygen concentration. 10. The method according to claim 8 and 9, characterized in that as Reed substrate a protein or a protein mixture is used. 11. The method according to claim 8 and 9, characterized in that as Reed substrate a lipid or a lipid mixture is used. 12. The method according to claim 8 and 9, characterized in that as Reed substrate a carbohydrate or a carbohydrate mixture is used. 13. The method according to claims 1 to 8 and one or more of the subsequent claims, characterized in that waste water quotas different composition from industrial, agricultural, biological or municipal areas merged, one after the other without Modification and calibration of the monitoring device and the neural Network are monitored and controlled. 14.Device for monitoring and influencing process flows of biological sewage treatment plants and treatment stages for the treatment of wastewater and wastewater mixtures of changing composition by determining the properties of the wastewater, before being introduced into the treatment plant, consisting of a monitoring tank with a mixing device for taking up the wastewater, the Activated sludge, the growth substrate and the oxygen, from measuring probes, a computer and dosing devices, the measuring probes ( 5 ) in the wastewater ( 1 ) measuring the chemical, biological or physical parameters in parallel and simultaneously, are connected to measuring data converters ( 9 ) in which the values determined during the parameter measurement are converted into electronic data records, the measurement data converters ( 9 ) are connected to the computer ( 11 ) for transmitting the data records via electronic interfaces ( 10 ), the computer ( 11 ) is connected to a neural network ( 18 ) a which generates output values from the data records, which are used to control the monitoring device ( 14 ) and to assess the quality of the wastewater, and the computer ( 11 ) for controlling ( 14 ) and repeating the monitoring processes several times with the measuring probes ( 5 ), the mixing device ( 7 ) and the metering devices ( 12 ) is electronically connected. 15. The apparatus according to claim 14, characterized in that at least two different measuring probes ( 5 ) are arranged on the container ( 6 ), with which the parameters are measured several times and the measured values thus determined after the conversion in the measuring data converter ( 9 ) as qualitatively different, however, time-dependent data records can be processed in the neural network.