AI Operator is based on the own proper patented technology of combination AI and thermodynamic
Thermodynamics is the basis of the product
Thermodynamic module is one of two key components of AI Operator. Basing of the AI Operator system on the thermodynamic modeling of the chemical unit is a guarantee of reliability of optimizational solutions which are confirmed by leading technologies of machine learning and artificial intelligence. During the implementation of AI Operator under the current project, the thermodynamic module is tuned up for the existing chemical unit in order to consider nuances of unit's apparatus implementation, catalytic system, aspects of process piping and specificity of unit control. This helps to precisely describe the parameters of the working regime and characteristics of products.
Thermodynamic module is used at the stage of quality analysis of historical data from the unit for the purpose to select the reference intervals for further machine learning. Simultaneously the thermodynamic module helps to fill in the gaps in the historical data, existing as a result of specificity of unit's DCS, rare laboratory analyses, instruments losses. The main purpose of the module is selection of the optimal regime of unit operation considering the current inlet parameters, equipment limitations, external requirements and factors.
Statistics – the breakthrough algorithms
The second key component of AI Operator is a statistic module, operating on the basis of artificial intelligence. Module is built on the data of the current unit, thus considering it's specific aspects. During the module learning, the selected historical data from the unit is used in combination with the data obtained from the created thermodynamic module. The key purpose of the machine learning module is determination of target indicators of unit operation, e.g. products quality parameters, basing on the current unit state, tracked by DCS.
Target parameters are determined in the real time mode without delays, what is of high importance for optimization of process unit operation. The artificial intelligence, basing on the historical data, predicts the values of the target parameters in the required time horizon and provides the optimal regime of unit operation with consideration of this prediction.
Combination of thermodynamics and statistics
AI Operator involves two modules which simultaneously operate together: thermodynamic and statistic modules. As a result all leading algorythms of artificial intelligence are applied in conjunction with reliable thermodynamic basis for development of a final optimizational decision. AI Operator, performs the role of unit optimizer, which considers all necessary and significant information for the process. Guides for the operator or the inputs for the DCS (depending on the AI Operator implementation) are developed considering feed type and its forecast, capabilities of unit equipment and its current load, market conjuncture, weather parameters, etc.
The combination of leading approaches helps to avoid the need to perform the factorial experiment at the commercial unit, thus to avoid economic losses. During the operation the system is re-studied on the continiously incoming data, what lets to provide the optimizational instructions of high accuracy. AI Operator may be implemented at the unit equipped only with the DCS, as well as at the unit equipped with DCS in combination with one or more APC modules.