WO2022184624A1

WO2022184624A1 - System, method and computer program for economically optimising a process in a production plant

Info

Publication number: WO2022184624A1
Application number: PCT/EP2022/054953
Authority: WO
Inventors: Carsten Andreas Klein; Christoph Haeusler; Wilfried Runde
Original assignee: Sms Group Gmbh
Priority date: 2021-03-01
Filing date: 2022-02-28
Publication date: 2022-09-09
Also published as: DE102021201887A1

Abstract

The invention relates to a system (1), method and computer program for economically optimising a process in a production plant (2), in particular a production plant (2) in the metallurgical industry, the non-ferric or steel industry or master alloy production, the production plant (2) comprising: a production planning system (3) for defining a production sequence for the production plant (2), in particular control data for an automation system (4); the automation system (4) for controlling production in the production plant (2), the automation system (4) preferably executing the production sequence defined by the production planning system (3); a state monitoring system (5) for sensing states of the production plant (2) and the components thereof; a maintenance planning system (6) for defining maintenance measures for the production plant (2); and a business planning system (7), associated with the production plant (2), for the economic management of production and maintenance in the production plant (2), characterised in that the system (1) comprises an optimiser (8), which links the data of the production planning system (3), of the automation system (4), of the state monitoring system (5) and of the maintenance planning system (6) with the data of the business planning system (7) in order to optimise the process in the production plant (2) economically.

Description

SYSTEM, METHOD AND COMPUTER PROGRAM FOR ECONOMIC OPTIMIZATION OF A PROCESS IN A PRODUCTION PLANT

The invention relates to a system for the economic optimization of a process in a production facility, in particular a production facility in the metal-producing industry, in the non-iron or steel industry or in the production of master alloys. The invention relates to a method and a computer program for the economic optimization of a process in a production plant, in particular a production plant in the metal-producing industry, the non-iron or steel industry or the production of master alloys. The production plant of the metal-producing industry is used in particular for the production of slabs, sheets, pipes, beams, rails, other profiles or strips and forged products and their preliminary products or for carrying out individual work steps within the scope of production. The production is planned with sufficient advance notice. The aim is to manufacture products with a certain product quality. At the same time, the production plant, especially the machines used, is subject to wear and tear and other failure mechanisms. Furthermore, suitable personnel is required for the manufacture of the products, as well as corresponding resources and energy sources.

It is known from the prior art to optimize the manufacturing process in a production plant in the metal-producing industry using systems that take over parts of the process monitoring and/or process automation. These include, for example, production planning systems, condition monitoring systems or quality management systems. These systems use, for example, measured values from the production plant or external data to orders that are required for the respective area of responsibility and evaluate them. At the same time, systems are used to optimize the maintenance and repair of the production plant. In this regard, it is known to depict economic aspects in the form of general rules, which then require prioritization of the individual process steps at the technical level. In addition, other maintenance and repair functions, such as material procurement and warehousing, are managed technically and commercially.

Furthermore, so-called ERP systems (Enterprise Resource Planning) are known from the prior art, which, for example, manage the commercial customer orders for the production plants. In an ERP system, delivery dates, prices, product specifications and

Delivery terms managed. From this commercial data, technical process steps must be defined for the operation of the production plant, which must be carried out in succession in order to manufacture a specified product. A production plan is created from all of these technical process steps and delivery dates can be determined and checked based on the production plan created. Due to the typical separation of technical planning from commercial/economic planning in production planning, resource purchasing, the procurement of spare parts, maintenance planning, order management to achieve certain product qualities, or the like, according to the state of the art only rough guide values for the economic Operation of the production plant specified. There is no precise knowledge of the costs and income associated with a decision. For this reason, a consistent, system-based mode of operation geared towards economic success has not been possible up to now. As a result, production facilities are operated in an economically suboptimal manner and operators cannot even provide information on the economic performance that could be achieved. For example, a good condition of the production plant is generally advantageous for product quality. Ensuring this requires the use of resources for maintenance and spare parts. On the other hand, there is a higher output, higher machine availability and possibly higher product prices. However, an economically sound optimization does not usually take place, especially not automatically and continuously.

In many cases, order management is also based on insufficient information. For example, an even utilization of all production lines (regardless of their added value) or a high proportion of high-margin products (without taking into account that this may result in costs elsewhere) or the highest possible throughput with fundamentally value-adding products (without an overview of whether higher profits could be achieved with less throughput with a different product mix).

A higher product quality can often be achieved by using high-quality (and therefore usually more expensive) input materials (e.g. ore, alloy, scrap). A well-founded optimization of how much additional costs for input materials are economically justified compared to higher output or higher product prices does not take place, especially not automatically for all products.

Furthermore, with regard to maintenance, the service life of components of the production plant can be influenced by the intensity of maintenance work. If there are several possible suppliers for components, these alternatives often vary in terms of the purchase price, the required maintenance measures and the susceptibility to wear. A consideration of the economically optimal maintenance concept, taking into account the service life, costs for maintenance, reliability of the components and Purchase prices do not take place, in particular not automatically for each component.

It is known from DE 10 2007 036 325 A1 to take into account the state of the production plant, in particular individual components of the production plant, during production planning. Furthermore, according to DE 10 2007 036 325 A1, a maintenance plan can be taken into account in the production planning. However, an economic optimization of the production process in the production plant does not take place, but maximum availability of the production plant is also turned off, regardless of the economic benefit.

EP 1 530 770 B1 discloses a device for optimizing individual process parameters of a production plant, taking into account a company and production planning system. In particular, according to EP 1 530 770 B1, an evaluation is made as to whether the financial outlay for the company and production planning system pays off through the optimization achieved, ie the return on investment achieved is determined. However, there is no economic optimization of the production process, only the effects of optimized process parameters on the effort involved are examined.

According to JP 2014109845 A, in the production planning for a production plant in the metal-producing industry, when determining a production sequence, the currently available orders and a forecast of orders expected in the future are taken into account in order to optimize the economic operation of the production plant. The optimization of the production sequence is therefore based on the prediction of future orders and there is no economic optimization of the current production process. WO 2018/145947 A1 discloses a method and system for planning production and/or maintenance in a plant in the metal-producing industry. The planning is based on status monitoring of the underlying automation system and also takes into account the data from a quality determination system for determining the product qualities achieved. Optimized production planning and/or maintenance for the production plant is derived from linking the status data with the product qualities achieved. The technical condition of the production plant is therefore monitored and taken into account in future production and/or maintenance. However, there is no targeted economic optimization of the production process within the production plant.

Proceeding from this prior art, the object of the invention is to optimize the profit that can be achieved with a production plant, in particular in the metal-producing industry, or to provide an estimate for the maximum profit that can be achieved.

The object is achieved according to the invention by a system for the economic optimization of a process in a production plant, in particular a production plant in the metal-producing industry, the non-iron or steel industry or the production of master alloys, the production plant comprising: a production planning system for determining a production sequence for the production plant, in particular of control data for an automation system, the automation system for controlling production in the

Production plant, wherein the automation system preferably executes the production sequence determined by the production planning system, a status monitoring system for recording the status of the production plant and its components, a maintenance planning system for determining maintenance measures for the production plant, a business planning system associated with the production plant for the business management of production and maintenance in the production plant, and a quality predictor, a condition predictor, a cost predictor, a process predictor, a resource predictor and/or a personnel predictor which is characterized by this that the system includes an optimizer that links the data from the production planning system, the automation system, the condition monitoring system and the maintenance planning system with the data from the corporate planning system in order to optimize the process in the production plant from an economic point of view, the optimizer using the predictors to predict the future plant states, future achievable product qualities, future costs or similar relevant properties, parameters or the like in the business Op optimization of the process in the production plant.

According to the invention, the technical parameters relevant to the process in the production plant, such as the production sequence, the condition of the production plant and its components and planned maintenance measures, are linked to the relevant business information from the corporate planning system in order to optimize the process in the production plant from a business management point of view. The process is, for example, a manufacturing process, maintenance process or the like. In contrast to the prior art, there is no technical optimization, such as maximizing the production volume or plant availability, but the economic optimum is determined, for example taking into account the necessary capital investment, including the negotiated delivery conditions Contractual penalties, storage costs, personnel costs, material costs, maintenance costs and the like.

According to the invention, all decisions regarding the process in the production plant are made in such a way that the added value of the

Plant operator is maximized. This is done by tracking all value streams of the production operation and evaluating and using this information to generate descriptive models for value creation or for the boundary conditions to be considered, in particular through the use of explicit optimization tools that maximize value creation. The core of the invention lies in using a technical system to record and take into account all aspects relevant to a business decision-making process in order to always be able to automatically determine the optimum for the entire business.

The traditionally separate worlds of the business control level and the subordinate technical levels are combined in the technical system according to the invention, so that business optimization becomes a technical problem which can be solved by the technical system according to the invention.

The enterprise planning system can be a direct part of the production plant or is assigned to it as an external component. According to a preferred variant of the invention, the system carries out the optimization repeatedly, in particular at regular intervals or after changes in parameters relevant to the optimization.

According to a variant of the invention, the system determines an economically optimized production sequence, economically optimized target values for the automation system, economically optimized maintenance measures and/or economically optimized administration measures. The result of this optimization is, for example, an economically optimized production sequence in connection with economically optimized maintenance measures, the optimized production sequence including the economically optimized set value specifications for the automation system. The administrative measures that have been optimized from an economic point of view relate, for example, to the purchase of materials and/or raw materials, storage, transport or the like. The economically optimized production sequence, economically optimized target values for the automation system, economically optimized maintenance measures and/or economically optimized administrative measures are preferably forwarded to the corresponding production planning system, automation system, maintenance planning system or business planning system.

In particular, the optimizer creates instructions for action regarding the production sequence, the maintenance and repair measures to be carried out, the acceptance of delivery orders, the initiation of orders for input materials, operating resources and/or spare parts, the creation of deployment plans for the staff, in particular the maintenance staff, or the like.

In a preferred variant of the invention, the system creates a personnel deployment plan for production and/or maintenance, a resource procurement plan, an energy procurement plan, a spare parts procurement plan, a maintenance plan or the like as part of the business optimization. The plans created are forwarded to the appropriate production planning system, maintenance planning system or corporate planning system for implementation. According to a variant of the invention, the system includes a quality catalogue, a process catalogue, a maintenance catalogue, a cost catalogue, a spare parts catalogue, a resource catalogue, a personnel catalogue, and/or a profit catalogue. The quality catalog contains, for example, all verifiable quality requirements that can be placed on products. In particular, the quality catalog contains the set of all producible products with their respective quality requirements. The process catalog lists process criteria that are relevant to product qualities, plant output, plant efficiency or the like. The process catalog contains, for example, process criteria that are required to achieve a specific quality requirement from the quality catalog. The maintenance catalog lists all maintenance measures that can be carried out, in particular the maintenance measures that are suitable or necessary to eliminate an existing or expected restriction. Furthermore, the maintenance catalog contains in particular the necessary spare parts, the necessary personnel requirements, the requirements for the production conditions, such as downtimes, as well as other necessary tools or the like with regard to the maintenance measures listed in the maintenance catalogue. The cost catalog in turn contains the current costs for resources, personnel deployment, spare parts and the like. If the costs can only be determined as a function of certain factors, then these factors, for example base costs, coefficients, etc., are stored in the catalog of costs. For example, the spare parts catalog contains a list of all mechanical, electrical, hydraulic or other components of the system. In particular, the spare parts catalog contains the current inventory as well as planned deliveries of spare parts and/or operating resources. Of the

Resource catalog lists in particular all input materials, production aids, consumables,

Energy sources, etc. on. The personnel catalog preferably contains all employees and, if necessary, known service providers, in particular with their formal qualifications, i.e. their skills from the activities to execute maintenance catalog and/or manufacturing process steps. Current prices and/or margins based on the current system analysis are stored in the profit catalogue. If the prices and/or margins can only be calculated from properties of the manufacturing process, then the corresponding relevant factors, coefficients or the like are stored. Using the catalogs mentioned and the information they contain, the process in the production plant can be optimized in a more targeted manner from an economic point of view. For this purpose, the information from the catalogs mentioned is linked to the products to be manufactured in the production plant and their quality requirements, and an economically optimized production sequence and, if necessary, economically optimized maintenance measures are determined.

In an advantageous variant of the invention, the system also includes a process monitor, a status monitor, a maintenance monitor, a planning monitor, a resource monitor, a quality monitor, a profit monitor and/or a personnel monitor. For example, the process monitor records process data, preferably all process data, from production and any accompanying data, such as e.g.

laboratory test results. The recorded data can be used for system analysis. The status monitor reflects the current and/or predicted status of the production plant according to all known evaluations and/or system analyses, preferably with regard to existing or expected restrictions, in particular based on entries from the process catalogue. The maintenance monitor indicates the current status of all maintenance measures that have already been planned. The planning monitor contains the current status of the planned production sequence. The current and already planned stock of input material is stored in the resource monitor, preferably including the already planned inflows of resources. The quality monitor contains the qualities of the individual products that have been achieved and/or that can be achieved in the future, in particular all quality data determined for the manufactured products. In the profit monitor, the value added for each type of product and each individual process step, in particular costs for poor quality can be derived for each process step or these are stored directly in the profit monitor. The personnel monitor reflects the current personnel availability, for example shift schedules, planned assignments by service providers, or the like.

According to a preferred variant of the invention, the predictions of the quality predictor, the status predictor, the cost predictor, the process predictor, the resource predictor and/or the personnel predictor and/or the business optimization of the optimizer are based on the process monitor, the status monitor, the maintenance monitor, the planning monitor, the resource monitor, the quality monitor, the profit monitor and/or the personnel monitor are continuously adjusted. In particular, the adjustment is made by means of machine learning from the process, quality, failure data and other data from the monitors mentioned, with the predictors (prediction models) or the algorithms of the optimizer continuously adapting or updating themselves.

According to the invention, the system comprises a quality predictor, a condition predictor, a cost predictor, a process predictor, a resource predictor and/or a personnel predictor. The quality predictor is designed to predict qualities that can be generated. The status predictor is designed to predict future plant statuses and the cost predictor to predict future costs of costly entities, in particular costs for input materials, energy, operating resources, spare parts and the like. The process predictor is designed to predict future process states. The resource predictor allows a prediction of the development of resources in the future and the personnel predictor allows a prediction of future personnel deployment. The aforementioned predictors preferably take into account the current planning. The predictors serve as prediction models for predicting future ones Plant conditions, product qualities that can be achieved in the future, future costs or the like. By means of the predictors or prediction models, the future plant states, product qualities that can be achieved in the future, future costs or similar relevant properties, parameters or the like are taken into account in the economic optimization of the process in the production plant. By predicting future properties, parameters or the like of the production plant or the process, the economic optimization is significantly improved again. The predictors enable an assessment of how the quality, the plant states, the costs for input materials, energy, operating resources, spare parts and the like, the process states, the process states and/or the personnel will behave in the future based on the current planning. On the basis of these states, properties, etc. projected into the future, the optimizer can optimize the process in the production plant from an economic point of view.

The cost predictor also takes into account, in particular, monetary effects with regard to various alternatives in the manufacturing process of a product, such as with regard to the process route (which process steps are passed through), the production lines passed through (if there are several comparable ones), and/or the exact design of a process characterized by its technological targets.

According to a variant of the invention, the quality predictor, the state predictor, the cost predictor, the process predictor, the

Resource predictor and / or the personal predictor each designed as a data model. The respective data model is preferably based on machine learning methods, in particular on artificial neural networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or non-linear regression models with or without regularization, support vector machines with linear, polynomial or other kernel functions, or the like.

In an expedient variant of the invention, the cost predictor is designed as a price catalog in which costs per consumption unit or per time unit are specified or from which the costs can be calculated according to a defined formula. In this way, the costs can be determined easily and quickly. Alternatively, the costs can also be determined on the basis of historical data from accounting or similar business systems.

According to a further variant of the invention, the system comprises a resource planner, a sequence planner, a maintenance planner, a spare parts planner and/or a personnel planner. The resource planner is used to plan the procurement of input materials, operating resources, energy and the like. The aim here is cost-optimal procurement, taking into account the existing production plan. In this context, cost-optimal takes into account, in particular, input material-dependent margins, taking into account potential effects of the input material or mixtures of input materials on the quality. The sequence planner is used to determine the production sequence of the existing orders. The planning of the production sequence can be done in different granularity/level of detail for different time horizons. The aim here is to maximize the added value for a given period of time. A predicted plant status is preferably taken into account, which in turn can be used for a comparison between the required product quality and the product quality that can be achieved according to the forecast. The sequence planner takes into account, for example, maintenance measures including the necessary parts and tools as well as the availability of resources and personnel. The aim is to maximize the profitability of plant operation. The maintenance planner is used to define maintenance measures, in particular the maintenance measures from the maintenance catalogue. These must meet the requirements of production planning when products are planned that, for example, require plant states according to the process catalogue, which are not fulfilled according to a current forecast. Measures from the maintenance catalogue, for example, are then to be taken in order to improve the condition of the system so that the requirements are met. This includes, for example, the exchange of tools such as work rolls, molds or the like. The spare parts planner is used to plan the procurement of spare parts and operating resources. Based on the current status, possibly including the inflows that have already been determined, as well as one according to the current one, if applicable

prediction of the expected consumption, the difference between the spare parts and operating resources required for the planned maintenance measures and the spare parts and operating resources required according to the spare parts catalog must be procured. The personnel planner is used to schedule personnel and, if necessary, to procure service providers. Sufficient personnel capacities with suitable roles must be provided for production and maintenance in accordance with the planned production sequence and planned maintenance measures. This is ensured by the personnel planner. The planners mentioned form the optimizer together or individually in order to optimize the process in the production plant from an economic point of view. The individual planners can be combined in an overall planner (overall optimizer) or designed as separate planners (partial optimizers), with the separate planners optimizing individual areas of the production plant from an economic point of view and, in combination, optimizing the process in the production plant from a financial point of view. In particular, the planners can be used to maximize the profitability of the production facility.

In an advantageous variant according to the invention, the system also includes an energy management system for further economic optimization of the process in the production plant. For this purpose, all of the optimizations mentioned are also taken into account with regard to optimization of the energy management examined. This includes the appropriate design of production planning, adapted to commercial contracts for energy supply, energetically optimal design of production or the like. Energy prices, especially electrical energy, fluctuate on several time scales. Due to climatic conditions, energy demand varies seasonally, as well as due to weather effects on scales of days, but also within a day, energy supply and demand fluctuate. Operators of industrial plants often have a mix of independently generated energy, energy storage and a mixture of long-term and short-term external procurement modes. On the one hand, the

When planning production, make sure that particularly energy-intensive or low-energy products are included in phases in which energy is forecast to be particularly cheap or expensive. In the economic evaluation, attention must be paid to the plant-internal generation or use of stored energy. Another degree of freedom is that, depending on the system layout, several processes often require electrical energy and the simultaneous planning of several demanding processes can lead to peak loads that can only be covered by purchasing energy at inflated prices. In this case, it would be the task of coupled energy and production planning to adapt the production sequence in such a way that no peak loads occur. Depending on the situation, the economic optimum can also consist of selling one's own long-term capacities on the energy market when energy prices are particularly high and not using them for production or even selling one's own generation or storage on the market.

According to an advantageous variant of the invention, the system includes a price generator for determining a minimum price for a product for economical production, a probable price for a product based on the current market situation and/or a highest possible, realistic price for a product. The price generator is used in particular for Determination of minimum prices for each product used to ensure economical production. Possible price reductions can be taken into account, for example caused by poor product quality, missed delivery dates or the like. Alternatively or additionally, a probable price for the product can be determined based on the current market situation and/or production utilization. A highest possible, realistic price for a product that can be achieved on the market could also be determined. The price generator takes into account the individual costs of manufacturing a product, the costs for the resources used and, if applicable, energy costs.

In an expedient variant of the invention, the price generator takes into account general costs of the production plant. The price generator thus enables a full cost view for the production of a specific product. For this purpose, overhead costs can also be allocated for parts of the total costs.

Since industrial production often takes place in production sequences, with similar products being produced within a sequence with the same components of a machine or system, and the interruptions in sequences leading to downtimes and the need to change production tools, heavily fragmented production represents a cost disadvantage. Depending on the order situation, however, it may not be possible to produce existing orders within a sequence, since not all product transitions are possible (e.g. due to geometry, materials or temperature requirements). An additional order that allows two existing sequences to be combined eliminates the costs associated with the downtime that would otherwise be required and is therefore economically attractive. The price generator can also take this price advantage into account and, if necessary, grant a discount to take account of the advantageous situation for the manufacturer. Furthermore, the price generator can be used to generate a recommendation for given or forecast achievable market prices as to which products should ideally be offered on the market, since the added value of different products on a given system is usually very different.

According to a variant of the invention, the price generator includes a static price catalog and/or an algorithm for dynamic price determination. For example, the price catalog is static and can be adjusted from the outside if necessary.

According to an expedient variant of the invention, the system includes a planning list containing information regarding the products to be produced, in particular their quality requirements. the

Quality requirements can be used, for example, as a reference to a

quality catalog be trained. The planning list is expediently part of the production planning system.

The optimizer expediently creates instructions for action regarding the production sequence, the maintenance and repair measures to be carried out, the acceptance of delivery orders, the initiation of orders for input materials, operating resources and/or spare parts, the creation of deployment plans for the personnel, in particular the

maintenance personnel, or the like.

In a variant according to the invention, the optimizer of the system according to the invention is based on reinforcement learning or operation research algorithms, in particular policy networks, neural networks, linear programs, convex optimization, mixed-integer optimization, or the like. The object is also achieved by a method for the economic optimization of a process in a production plant, in particular a production plant in the metal-producing industry, the non-iron or steel industry or the production of master alloys, the production plant comprising: a production planning system for determining a production sequence for the production plant, in particular control data for an automation system, the automation system for controlling production in the production plant, with the automation system preferably executing the production sequence determined by the production planning system, a status monitoring system for recording statuses of the production plant and its components, a maintenance planning system for determining maintenance measures for the production plant, a business planning system associated with the production plant for the economic management of production and Maintenance in the production plant, and a quality predictor, a condition predictor, a cost predictor, a process predictor, a resource predictor and/or a personnel predictor, the method linking the data from the production planning system, the automation system, the condition monitoring system and the maintenance planning system with the data from the enterprise planning system in order to to optimize the process in the production plant from an economic point of view, with the optimization using the predictors taking into account the future plant states, product qualities that can be achieved in the future, future costs or similar relevant properties, parameters or the like in the economic optimization of the process in the production plant. According to the invention, the technical parameters relevant to the process in the production plant, such as the production sequence, the condition of the production plant and its components and planned maintenance measures, are linked to the relevant business information from the corporate planning system in order to optimize the process in the production plant from a business management point of view. In contrast to the state of the art, there is no technical optimization, such as maximizing the production volume or plant availability, but the economic optimum is determined, for example taking into account the necessary capital investment, the negotiated delivery conditions including contractual penalties, storage costs, personnel costs, material costs, maintenance costs and the same.

The process to be optimized is, for example, a manufacturing process, maintenance process or the like.

In a preferred variant of the invention, the method carries out the optimization repeatedly, in particular at regular intervals or after changes in parameters relevant to the optimization.

According to the invention, all decisions regarding the process in the production plant are made in such a way that the added value of the plant operator is maximized. This is done by tracking all value streams of the production operation and evaluating and using this information to generate descriptive models for value creation or for the boundary conditions to be considered, in particular through the use of explicit optimization tools that maximize value creation. The core of the invention lies in using a technical method to record and take into account all aspects relevant to a business decision-making process in order to always be able to automatically determine the optimum for the entire business. The traditionally separate worlds of the business control level and the subordinate technical levels are combined in the technical method according to the invention, so that the business optimization becomes a technical problem which can be solved by the technical method according to the invention.

According to an advantageous variant of the invention, the method determines an economically optimized production sequence, economically optimized target values for the automation system, economically optimized maintenance measures and/or economically optimized administration measures. The result of this method step is, for example, an economically optimized production sequence in connection with economically optimized maintenance measures, the optimized production sequence including the economically optimized setpoint specifications for the automation system. The administrative measures that have been optimized from an economic point of view relate, for example, to the purchase of materials and/or raw materials, storage, transport or the like. The economically optimized production sequence, economically optimized target values for the automation system, economically optimized maintenance measures and/or economically optimized administrative measures are preferably forwarded to the corresponding production planning system, automation system, maintenance planning system or company planning system.

In an expedient variant of the invention, the method creates a personnel deployment plan for production and/or maintenance, a resource procurement plan, an energy procurement plan, a spare parts procurement plan, a maintenance plan or the like as part of the business optimization. The plans drawn up are sent to the appropriate production planning system, maintenance planning system or business planning system forwarded.

According to a variant of the invention, the method includes the creation and / or management of a quality catalog, a process catalog, a

maintenance catalogue, a cost catalogue, a spare parts catalogue, a resource catalogue, a personnel catalogue, and/or a profit catalogue. The quality catalog contains, for example, all verifiable quality requirements that can be placed on products. In particular, the quality catalog contains the set of all producible products with their respective ones

quality requirements. The process catalog lists process criteria that are relevant to product qualities, plant output, plant efficiency, or the like. The process catalog contains, for example, process criteria that are required to achieve a specific quality requirement from the quality catalog. The maintenance catalog lists all executables

Maintenance measures, in particular the maintenance measures that are suitable or necessary to remedy an existing or expected restriction. Furthermore, the maintenance catalog contains in particular the necessary spare parts, the necessary personnel requirements, the requirements for the production conditions, such as downtimes, as well as other necessary tools or the like with regard to the maintenance measures listed in the maintenance catalogue. The cost catalog in turn contains the current costs for resources, personnel deployment, spare parts and the like. If the costs can only be determined as a function of certain factors, then these factors, for example basic costs, coefficients, etc. are included in the

Cost catalog saved. For example, the spare parts catalog contains a list of all mechanical, electrical, hydraulic or other components of the system. In particular, the spare parts catalog contains the current inventory as well as planned deliveries of spare parts and/or operating resources. The resource catalog lists in particular all input materials, production aids, consumables, Energy sources, etc. on. The personnel catalog preferably contains all employees and possibly known service providers, in particular with their formal qualifications, that is to say their abilities to carry out activities from the maintenance catalog and/or production process steps. Current prices and/or margins based on the current system analysis are stored in the profit catalogue. If the prices and/or margins are only to be calculated from properties of the production process, then the corresponding relevant factors, coefficients or the like are stored. Using the catalogs mentioned and the information they contain, the process in the production plant can be optimized in a more targeted manner from an economic point of view. For this purpose, the information from the catalogs mentioned is linked to the products to be manufactured in the production plant and their quality requirements, and an economically optimized production sequence and, if necessary, economically optimized maintenance measures are determined.

According to an advantageous variant of the invention, the method includes creating and/or managing a process monitor, a status monitor, a maintenance monitor, a planning monitor, a resource monitor, a quality monitor, a profit monitor and/or a personnel monitor. The process monitor records, for example, process data, preferably all process data, from production and any accompanying data, such as laboratory test results. The recorded data can be used for system analysis. The status monitor reflects the current and/or predicted status of the production plant according to all known evaluations and/or system analyses, preferably with regard to existing or expected restrictions, in particular based on entries from the process catalogue. The maintenance monitor indicates the current status of all maintenance measures that have already been planned. The planning monitor contains the current status of the planned production sequence. The current and already planned stock of input material is stored in the resource monitor, preferably including the already planned inflows of resources. In the quality monitor are contain the qualities of the individual products that have been achieved and/or that can be achieved in the future, in particular all quality data determined for the manufactured products. The profit monitor contains the added value for each type of product and each individual process step. In particular, costs for poor quality can be derived for each process step or these are stored directly in the profit monitor. The personnel monitor reflects the current personnel availability, for example shift schedules, planned assignments by service providers, or the like. According to a preferred variant of the invention, the predictions of the quality predictor, the status predictor, the cost predictor, the process predictor, the resource predictor and/or the personnel predictor and/or the business optimization of the optimizer are based on the process monitor, the status monitor, the maintenance monitor, the planning monitor, the resource monitor, the quality monitor, the profit monitor and/or the personnel monitor are continuously adjusted. In particular, the adjustment is made by means of machine learning from the process, quality, failure data and other data from the monitors mentioned, with the predictors (prediction models) or the algorithms of the optimizer continuously adapting or updating themselves.

In a preferred variant according to the invention, the method includes the creation and/or management of a quality predictor, a status predictor, a cost predictor, a process predictor, a resource predictor and/or a personnel predictor. The quality predictor is designed to predict qualities that can be generated. The status predictor is designed to predict future plant statuses and the cost predictor to predict future costs of costly entities, in particular costs for input materials, energy, operating resources, spare parts and the like. The process predictor is designed to predict future process states. The resource predictor allows a prediction about the Development of resources in the future and the personnel predictor allows a prediction of future personnel deployment. The aforementioned predictors preferably take into account the current planning. The predictors serve as prediction models for predicting future plant states, product qualities that can be achieved in the future, future costs or the like. By means of the predictors or prediction models, the future plant states, product qualities that can be achieved in the future, future costs or similar relevant properties, parameters or the like are taken into account in the economic optimization of the process in the production plant. By predicting future properties, parameters or the like of the production plant or the manufacturing process, the economic optimization is significantly improved again. The predictors enable an assessment of how the quality, the plant states, the costs for input materials, energy, operating resources, spare parts and the like, the process states, the process states and/or the personnel will behave in the future based on the current planning. On the basis of these states, properties, etc. projected into the future, the optimizer can optimize the process in the production plant from an economic point of view. The cost predictor also takes into account, in particular, monetary effects with regard to various alternatives in the manufacturing process of a product, such as with regard to the process route (which process steps are passed through), the production lines passed through (if there are several comparable ones), and/or the exact design of a process characterized by its technological targets.

According to an expedient variant of the method according to the invention, the quality predictor, the condition predictor, the cost predictor, the process predictor, the resource predictor and/or the personnel predictor are each based on a data model. The respective data model is preferably based on machine learning methods, in particular on artificial neural ones Networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or non-linear regression models with or without regularization, support vector machines with linear, polynomial or other kernel functions, or the like.

In an expedient variant of the invention, the cost predictor is designed as a price catalog in which costs per consumption unit or per time unit are specified or from which the costs can be calculated according to a defined formula. In this way, the costs can be determined easily and quickly. Alternatively, the costs can also be increased

be determined on the basis of historical data from accounting or similar business systems.

According to an advantageous variant, the method according to the invention also includes the creation and/or management of a resource planner, a sequence planner, a maintenance planner, a spare parts planner and/or a personnel planner. The resource planner is used to plan the procurement of input materials, operating resources, energy and the like. The aim here is cost-optimal procurement, taking into account the existing production plan. In this context, cost-optimal takes into account, in particular, input material-dependent margins, taking into account potential effects of the input material or mixtures of input materials on the quality. The sequence planner is used to determine the production sequence of the existing orders. The planning of the production sequence can be done in different granularity/level of detail for different time horizons. The aim here is to maximize the added value for a given period of time. A predicted plant status is preferably taken into account, which in turn can be used for a comparison between the required product quality and the product quality that can be achieved according to the forecast. The sequence planner takes into account, for example, maintenance measures including the necessary parts and tools as well as resource and staff availability. The aim is to maximize the profitability of plant operation. The maintenance planner is used to define maintenance measures, in particular the maintenance measures from the maintenance catalogue. These must meet the requirements from production planning if products are planned that, for example, according to

Process catalogue, requiring plant states that are not fulfilled according to a current forecast. There are then, for example, measures from the

Carry out maintenance catalog to improve the condition of the plant accordingly, so that the requirements are met. This includes e.g. the exchange of tools such as work rolls, molds or the like. The spare parts planner is used to plan the procurement of spare parts and

operating tools. Based on the current status, possibly including the inflows that have already been determined, as well as one according to the current one, if applicable

prediction of the expected consumption, the difference between the spare parts and operating resources required for the planned maintenance measures and the spare parts and operating resources required according to the spare parts catalog must be procured. The personnel planner is used to schedule personnel and, if necessary, to procure service providers. Sufficient personnel capacities with suitable roles must be provided for production and maintenance in accordance with the planned production sequence and planned maintenance measures. This is ensured by the personnel planner. The planners mentioned form the optimizer together or individually in order to optimize the process in the production plant from an economic point of view. The individual planners can be combined in an overall planner (overall optimizer) or designed as separate planners (partial optimizers), with the separate planners optimizing individual areas of the production plant from an economic point of view and, in combination, optimizing the process in the production plant from a financial point of view. In particular, the planners can be used to maximize the profitability of the production facility. In an advantageous variant according to the invention, the method also includes the step of energy management for further economic optimization of the process in the production plant. For this purpose, all the optimizations mentioned are also examined with regard to optimization of the energy management. This includes the appropriate design of production planning, adapted to commercial contracts for energy supply, energetically optimal design of production or the like. Energy prices, especially electrical energy, fluctuate on several time scales. Due to climatic conditions, energy demand varies seasonally, as well as due to weather effects on scales of days, but also within a day, energy supply and demand fluctuate. Operators of industrial plants often have a mix of independently generated energy, energy storage and a mixture of long-term and short-term external procurement modes. On the one hand, production planning can now ensure that particularly energy-intensive or low-energy products are planned in phases in which energy is forecast to be particularly cheap or expensive. In the economic evaluation, attention must be paid to the plant-internal generation or use of stored energy. Another degree of freedom is that, depending on the system layout, several processes often require electrical energy and the simultaneous planning of several demanding processes can lead to peak loads that can only be covered by purchasing energy at inflated prices. In this case, it would be the task of coupled energy and production planning to adapt the production sequence in such a way that no peak loads occur. Depending on the situation, the economic optimum can also consist of selling one's own long-term capacities on the energy market when energy prices are particularly high and not using them for production or even selling one's own generation or storage on the market. According to a preferred variant of the invention, the method includes determining a minimum price for a product for economical production, a probable price for a product based on the current market situation and/or a highest possible, realistic price for a product. The minimum price is determined in particular for each product to be manufactured in order to ensure economical production. Possible price reductions can be taken into account, for example caused by poor product quality, missed delivery dates or the like. The probable price for the product is based in particular on the current market situation and/or production capacity. For example, the highest possible, realistic price is independent of the current market situation and also takes into account probable future market situations. The method takes into account the individual costs of manufacturing a product, the costs for the resources used and, if applicable, energy costs.

In an expedient variant of the invention, the minimum price, the probable price and/or the highest possible, realistic price takes into account the general costs of the production facility. The specific minimum price is therefore related to a full cost consideration for the production of a specific product. For this purpose, overhead costs can also be allocated for parts of the total costs.

According to a variant according to the invention, the determination of the minimum price, the probable price and/or the highest possible realistic price is based on a static price catalog and/or an algorithm for dynamic price determination. For example, the price catalog is static and can be adjusted from the outside if necessary.

According to an expedient variant of the invention, the method includes the creation and/or management of a planning list comprising information regarding the products that are to be produced, in particular their products quality requirements. The quality requirements can, for example, be in the form of a reference to a quality catalogue. The planning list is expediently part of the production planning system. The method expediently creates instructions for action regarding the production sequence, the maintenance and repair measures to be carried out, the acceptance of delivery orders, the initiation of orders for input materials, operating resources and/or spare parts, the creation of deployment plans for the personnel, in particular the maintenance personnel, or the like.

In a variant according to the invention, the method is based on reinforcement learning or operation research algorithms, in particular policy networks, neural networks, linear programs, convex optimization, mixed-integer optimization, or the like.

The object is also achieved by a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to the invention, in particular comprising instructions which cause the system according to the invention to execute the method according to the invention.

The computer program is designed in particular as a software module, which is located between levels 0-3 and 4 of the automation pyramid, i.e. between the automation of the manufacturing process in the production location (level 0-3) and the enterprise planning system (Enterprise Resource Planning ERP system at level 4 ).

The economic optimization of a process in a production plant, in particular a production plant in the metal-producing industry, the non-iron or steel industry or the production of master alloys, using the system according to the invention, the method according to the invention and/or the computer program according to the invention takes place according to an advantageous variant at specific time intervals and/or after specific events have occurred. The situation is constantly changing due to the ongoing production process, new customer orders, unforeseeable changes in plant availability, the product quality achieved, etc. In order to always achieve a situationally economically optimal mode of operation, it is therefore necessary to constantly adapt the various operational plans to the changing boundary conditions. It is therefore necessary to implement the holistic, economically optimized mode of operation through repeated planning steps.

If the business optimization of the process in the production plant is carried out by an overall planner (overall optimizer), who simultaneously takes over the business optimization of all operational processes to be planned, the business optimization is preferably carried out according to a fixed time pattern, for example once a day, per hour, or the like . If the economic optimization of the process in the production plant is carried out by several separate planners (partial optimizers), who each plan different parts of the operational processes (e.g. the deployment of personnel, maintenance, procurement of resources, etc.) and jointly optimize the process in the production plant, so the separate planners are preferably executed at different times, e.g. the personnel planner weekly, the resource procurement daily, etc.

The overall planner (overall optimizer) and/or the separate planners

(partial optimizers) cannot be executed or not only at fixed times, but also at times when certain events occur, such as the placement of a new customer order, the completion of a product piece or a message from the condition monitor indicating a continuation of the Production in the currently planned form is no longer allowed. If several individual separate planners are used, various triggering mechanisms can also come into play for them, including the completion of a new plan of a planning component as a trigger for another planning component.

The updating of predictors, catalogues, etc. can be done in the same way. Since, according to the invention, a model of the production plant that is as up-to-date as possible is to be used for planning, it is necessary to adapt the predictors in particular, but also the catalogs, to the current operational experiences insofar as these are recorded systemically. Here, too, a fixed time frame can be used (e.g. once a week) for all tools (predictors, catalogues, etc.). Alternatively, different but fixed times for updating can be used for subsets of the tools.

However, events can also be defined for all or part of the tools (predictors, catalogues, etc.), which are used as triggers for updating the tools, such as an incident for updating the condition predictor, a defective product for the Quality predictor or underrun of resource stocks for resource predictor update.

The implementation of the optimized mode of operation at the production plant is carried out by the personnel who are also deployed, but also in an essential way by the plant automation (automation system). It is the task of the quality predictor, in the course of determining the product quality that can be achieved over time, to define boundary conditions for the technical processes that ensure the achievement of desired quality characteristics, minimize the deviation from quality-relevant target values or the probability of missing limit values for quality characteristics minimize. This is done both by defining technological target values for the individual process steps as far as the automation system is able to use them directly. It can also be metadata for automation, such as special parameters that are required as part of a control or regulation and the product and

be defined specifically for the process state. In a preferred embodiment, the information required for this is determined, among other things, by machine learning of systematic relationships between production parameters (recorded in the process monitor) and the product quality achieved (recorded in the quality monitor). In a preferred embodiment, the actual target specifications for the system automation can then be determined by optimization algorithms, with the above-mentioned criteria as the goal. In a further embodiment, attention can already be paid here to the fact that the choice of process parameters also affects the state of the plant and thus the long-term

performance of the system or interactions with the choice of raw materials are possible. The system automation is enabled and authorized to implement the production method defined in this way, so that the

Subject to unforeseeable events and deviations between the model image of the system and the processes taking place and the real production situation that cannot be ruled out, production conditions largely correspond to those that were also expected during planning and thus, subject to the restrictions mentioned, the product quality achieved essentially corresponds to that announced during planning quality corresponds. Systematic deviations can be detected via the adaptation processes as part of the method, for example by the machine learning processes already mentioned, and can be used for future planning tasks to improve the accuracy of the prediction. The invention is explained in more detail below with reference to exemplary embodiments illustrated in the figures. Show it: 1 shows a schematic view of a system according to the invention for the economic optimization of a process in a production plant,

2 further details of an optimizer of a system according to the invention,

Fig. 3a - 31 examples of economic optimization of processes in a production plant, and

4 shows an example linking of components of a system according to the invention for the economic optimization of a process in a production plant,

5 shows an example linking of components of a system according to the invention for creating forecasts,

6 shows an exemplary linking of components of a system for resource planning according to the invention,

7 shows an example linking of components of a system according to the invention for procuring spare parts, FIG. 8 shows an example linking of components of a system according to the invention for procuring spare parts with personnel planning,

9 shows a further exemplary linking of components of a system according to the invention for the economic optimization of a process in a production plant, and Fig. 10 - 13 exemplary time sequences for the economic optimization of a process in a production plant.

1 shows a schematic view of a first embodiment of a system 1 according to the invention for the economic optimization of a process in a production plant 2, in particular a production plant 2 in the metal-producing industry, the non-ferrous or steel industry or master alloy production. The production plant 2 from Fig. 1 comprises a production planning system 3 for determining a production sequence for the production plant 2, in particular control data for an automation system 4. The automation system 4 of the production plant 2 from Fig. 1 is used to control production in the production plant 2, the Automation system 4 preferably executes the production sequence determined by the production planning system 3.

The production plant 2 also includes a status monitoring system 5 for detecting the status of the production plant 2 and its components, and a maintenance planning system 6 for determining maintenance measures for the production plant 2.

The production plant 2 from FIG. 1 also includes a business planning system 7 assigned to the production plant 2 for the business management of production and maintenance in the production plant 2. In the exemplary embodiment from FIG. 1, the business planning system 7 is designed as an external component, but it is also possible in principle that the business planning system 7 is an integral part of the production plant 2. The system 1 according to the invention is characterized in that it includes an optimizer 8 . The optimizer 8 links the data of the production planning system 3, the automation system 4, the

Condition monitoring system 5 and the maintenance planning system 6 with the data from the business planning system 7 in order to optimize the process in the production plant 2 from an economic point of view.

The system 1 from FIG. 1 determines in particular an economically optimized production sequence, economically optimized target values for the automation system 4 , economically optimized maintenance measures and/or economically optimized administration measures. As part of the business optimization, the system 1 creates, for example, a personnel deployment plan for production and/or maintenance, a resource procurement plan, an energy procurement plan, a spare parts procurement plan, a maintenance plan or the like.

The optimizer 8 is based, for example, on reinforcement learning algorithms or operation research algorithms, in particular policy networks, neural networks, linear programs, convex optimization, mixed integers

optimization, or something like that.

2 shows further details of an optimizer 8 of a system according to the invention. The optimizer 8 includes a number of components which can be divided into the categories catalogs, monitors, planners and predictors, for example.

The catalog category includes, for example, a quality catalog 9, a process catalog 10, a maintenance catalog 11, a cost catalog 12 and a planning list 26. The quality catalog 9 contains, for example, the quantity of all producible products with their respective quality requirements. In which Process catalog 10 lists process criteria that are required to achieve a specific quality requirement from the quality catalog 9 . The maintenance catalog 11 lists maintenance measures that are suitable or necessary to eliminate an existing or expected restriction. Furthermore, the maintenance catalog 11 contains in particular the necessary spare parts, the necessary personnel requirements, the requirements for the production conditions, such as downtimes, and other necessary tools or the like with regard to the maintenance measures listed in the maintenance catalog 11. The cost catalog 12 in turn contains the current costs for resources, personnel deployment, spare parts and the like. the

Planning list 26 includes, for example, information regarding the to

Products to be produced, in particular their quality requirements. The quality requirements can be used, for example, as a reference to a

Quality catalog 9 trained. The process in production plant 2 can be optimized in a more targeted manner from an economic point of view using the catalogs mentioned and the information contained therein. For this purpose, the information from the named catalogs is linked to the products to be manufactured in the production facility 2 and their quality requirements, and an economically optimized production sequence and possibly economically optimized maintenance measures are determined. the

Planning list 26 includes, in particular, information regarding the to

Products to be produced, such as their quality requirements. The Catalogs category also includes a

Spare parts catalog 27, the spare parts catalog 27 containing, for example, the current stock and planned supplies of spare parts and/or resources.

The monitors category includes, for example, a process monitor 13, a status monitor 14, a maintenance monitor 15 and a quality monitor 16. The process monitor 13 contains, for example, process data from production and any accompanying data, such as laboratory test results. The status monitor 14 contains the current and/or forecast State of the production plant 2 with regard to existing or expected restrictions, in particular based on entries from the process catalog 10. The maintenance monitor 15 contains information on all maintenance measures that have already been planned. The quality monitor 16 contains the qualities of the individual products that have been achieved and/or that can be achieved in the future.

The monitors category may also include a planning monitor 28, a resource monitor 29 and/or a profit monitor 30. The planning monitor 28 contains in particular the current status of the planned production sequence. In the resource monitor 29, for example, the current and already planned inventory of input material is stored. The value added for each type of product and each individual process step is usefully contained in the profit monitor 30, in particular costs for poor quality can also be derived for each process step or these are stored directly in the profit monitor 30.

The planner category includes, for example, a resource planner 20, a sequence planner 21, a maintenance planner 22, a spare parts planner 23, a personnel planner 24 and a price generator 25. The resource planner 20 is used to plan the procurement of input materials, operating resources, energy and the like. The aim here is cost-optimal procurement, taking into account the existing production plan. In this context, cost-optimal takes into account, in particular, input material-dependent margins, taking into account potential effects of the input material or mixtures of input materials on the quality. The sequence planner 21 is used to determine the production sequence of the existing orders. The planning of the production sequence can be done in different granularity/level of detail for different time horizons. The aim here is to maximize the added value for a given period of time. A predicted system status is preferably taken into account. The maintenance scheduler 22 is used to define maintenance measures. These must meet the requirements of production planning if products are planned that, for example, according to process catalog 10, require system states that are not fulfilled according to a current forecast. Then, for example, measures from the maintenance catalog 11 are to be taken in order to improve the condition of the system accordingly, so that the requirements are met. This includes, for example, the exchange of tools such as work rolls, molds or the like. The spare parts planner 23 is used to plan the procurement of spare parts and operating resources. Based on the current status, possibly including the inflows that have already been determined, the difference between the spare parts and operating aids required for planned maintenance measures and the spare parts and operating aids required according to a possible spare parts catalog 27 must be procured. The personnel planner 24 is used for scheduling personnel. Sufficient personnel capacities with suitable roles must be provided for production and maintenance in accordance with the planned production sequence and planned maintenance measures. The personnel planner 24 ensures this. The price generator 25 is used to determine a minimum price for a product for economical production, a probable price for a product based on the current market situation and/or a highest possible, realistic price for a product. The price generator 25 is used in particular to determine minimum prices for each product in order to ensure economical production. In particular, the price generator 25 takes into account the general costs of the production plant 25. Possible discounts can also be taken into account, for example caused by poor product quality, missed delivery dates or the like. The price generator 25 can also comprise a static price catalog and/or an algorithm for dynamic price determination. A probable price for the product can preferably also be determined based on the current market situation and/or production utilization. Also he can Price generator, for example, determine a highest possible, realistic price.

The category of predictors includes, for example, a quality predictor 17, a state predictor 18, a cost predictor 19 and/or a process predictor 31. The quality predictor 17 is designed to predict qualities that can be produced. The status predictor 18 is designed to predict future plant statuses and the cost predictor 19 to predict future costs for input materials, energy, operating resources, spare parts and the like. The process predictor 31 is designed to predict future process states. The predictors serve as prediction models for predicting future plant states, product qualities that can be achieved in the future, future costs or the like. The future plant states, product qualities that can be achieved in the future, future costs or similar relevant properties, parameters or the like are taken into account in the economic optimization of the process in the production plant 2 by means of the predictors or prediction models. By predicting future properties, parameters or the like of the production facility 2 or the manufacturing process, the economic optimization is again significantly improved.

The cost predictor 19 also takes into account, in particular, monetary effects with regard to various alternatives in the manufacturing process of a product, such as with regard to the process route (which process steps are run through), the production lines run through (if there are several comparable ones), and/or the exact design of a process characterized through its technological target specifications.

According to a preferred variant of the invention, the quality predictor 17, the state predictor 18 and the cost predictor 19 are each designed as a data model. The respective data model is based, for example, on methods from the machine learning, in particular on artificial neural networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or nonlinear regression models with or without regularization, support vector machines with linear, polynomial or other kernel functions, or the like.

The cost predictor 19 can also be in the form of a price catalogue, in which costs per consumption unit or per time unit are specified, or from which the costs can be calculated according to a defined formula. Alternatively, the costs can also be determined on the basis of historical data from accounting or similar business systems.

In the figures 3a to 31 examples of economic optimization of processes in a production plant are shown symbolically.

3a shows symbolically how future states of the production plant 2 can be predicted. The knowledge of future states of the production plant 2 allows the manufacturing process in the production plant 2 to be economically optimized since the expected states of the production plant 2 can be reacted to at an early stage. To predict future states of the production plant 2, the optimizer 8 links the data from the planning list 26, the state monitor 14, the maintenance monitor 13 and the planning monitor 28. The linked data are supplied to the state predictor 18, which predicts future states of the production plant 2. This prediction of future states of the production facility 2 is made available to the state monitor 14 .

3b symbolically shows the prediction of product qualities that can be achieved in the future, which is an essential factor for optimizing the processes in a production plant 2 from an economic point of view. For this purpose, the data of the process catalog 10 and the quality catalog are provided to the quality predictor 17 9 transmitted together with the data from the status monitor 14. From this data, the quality predictor 17 can predict product qualities that can be achieved in the future, which can be made available to the quality monitor 16 .

3c symbolizes the optimization of the manufacturing process in a production plant 2 by planning future production. For this purpose, the data of the planning list 26 are transmitted together with the data of the maintenance monitor 15 and the quality monitor 16 to the sequence planner 21, which determines the future production from the transmitted data and thereby optimizes it. The results are preferably transmitted to the planning monitor 28 which is designed to monitor the production in the production plant 2 . In Fig. 3d the planning of the future procurement of input material for the manufacturing process in a production plant is shown symbolically. By optimizing the future procurement of input material, the manufacturing process can be economically optimized. For this purpose, the data from the cost catalogue, the planning monitor 28 and the resource monitor 29 are transmitted together with the forecast from the cost predictor 19 to the resource planner 20, which uses the transmitted data to determine an economically optimized procurement of materials for the future. The result is in turn made available to the resource monitor 29 .

3e symbolically shows the determination of economically optimized product prices, in particular to ensure profitable production. In this regard, the price generator 25 uses the data from the cost catalog 12, the cost predictor 19 and the planning monitor 28 to determine a minimum price for the products to be manufactured, a probable price for the product based on the current market situation and/or a maximum realistic price. The price generator 25 takes into account the general costs of the production plant 2, for example. The minimum prices determined by the price generator 25 are transmitted to the profit monitor 30, which can carry out a business optimization of the product prices on the basis of the data provided.

The updating of a maintenance plan for a production plant 2 is shown symbolically in FIG. 3f. The maintenance work in the production plant 2 can thus be optimized from an economic point of view. To do this, the maintenance planner 22 uses the data from the maintenance catalogue, the process catalogue, the

cost catalog and the spare parts catalogue. The maintenance planner 22 also receives data from the status monitor 14, maintenance monitor 15, quality monitor 16, planning monitor 28 and profit monitor 30. The updated maintenance plan is transmitted from the maintenance planner 22 to the maintenance monitor 15, for example to monitor the execution of the updated maintenance plan.

3g symbolically shows the updating of a personnel deployment plan for the execution of processes in a production facility 2. For the economic optimization of personnel deployment, the personnel planner 24 receives data from the maintenance catalog 11, cost catalog 12 and

Spare parts catalog 27, as well as from the maintenance monitor 15 and planning monitor 28. The optimized personnel deployment plan can optionally be sent to the profit monitor 30 for further processing. An economically optimized procurement of spare parts is shown symbolically in FIG. 3h. For this purpose, the data from the cost catalog 12 and

Spare parts catalog 27 transmitted to the spare parts planner 23 together with the data from the maintenance monitor 15 and profit monitor 30 . The spare parts planner 23 determines the economically optimized procurement of spare parts and makes the result available to the spare parts catalog 27 . 3i symbolically shows the learning of economically optimized processes to achieve defined quality features. In this regard, the data from the condition monitor 14, maintenance monitor 15, process monitor 13 and quality monitor 16 are transmitted to the process predictor 31. The process predictor 31 is to adapt the process catalog 10 according to the optimization.

In Fig. 3j the prediction of producible product qualities is shown. To do this, the quality predictor 17 uses the data from the status monitor 14, maintenance monitor 15, process monitor 13 and quality monitor 16.

The learning of optimized maintenance measures to improve the condition of the production plant 2 is shown in FIG. 3k. To do this, the process predictor 31 uses the data from the status monitor 14 and maintenance monitor 15 in order to update the maintenance catalog 11 .

Fig. 31 symbolically shows the learning of holistic manufacturing costs, margins and costs for defective quality. To do this, the cost predictor 19 uses the data from the cost catalog 12 and replacement catalog 13, together with the data from the condition monitor 14, maintenance monitor 15, process monitor 13 and quality monitor 16. The cost predictor 19 can update the cost catalog 12 if necessary.

The above examples are not mutually exclusive and can be combined with one another as desired.

4 shows an exemplary linking of components of a system 1 according to the invention for the economic optimization of a process in a production plant 2, in particular a production plant 2 in the metal-producing industry, the non-iron or steel industry or the production of master alloys. The production plant 2 comprises the following subsystems: a production planning system 3 for determining a production sequence for the production plant 2, in particular control data for an automation system 4, the automation system 4 for controlling production in the production plant 2, with the automation system 4 preferably using the production planning system 3 executes a specific production sequence, a status monitoring system 5 for detecting statuses of the production plant 2 and its components, a maintenance planning system 6 for determining maintenance measures for the production plant 2, and a corporate planning system 7 associated with the production plant 2 for the business management of production and maintenance in the production plant 2.

According to the invention, the system 1 includes an optimizer 8, which links the data from the production planning system 3, the automation system 4, the condition monitoring system 5 and the maintenance planning system 6 with the data from the corporate planning system 7 in order to optimize the process in the production plant 2 from an economic point of view.

The system 1 determines, in particular, an economically optimized production sequence, economically optimized target values for the automation system 4, economically optimized maintenance measures and/or economically optimized administration measures. As part of the business optimization, the system 1 according to the invention creates, for example, a personnel deployment plan for production and/or maintenance, a resource procurement plan, an energy procurement plan, a spare parts procurement plan, a maintenance plan or the like.

According to Fig. 4, the system 1 according to the invention comprises a quality catalog 9, a process catalog 10, a maintenance catalog 11, a cost catalog 12, a spare parts catalog 27, a resource catalog 32, a personnel catalog 33 and a profit catalog 34.

Furthermore, the system 1 according to the invention according to FIG. 4 comprises a process monitor 13, a status monitor 14, a maintenance monitor 15, a planning monitor 28, a resource monitor 29, a quality monitor 16, a profit monitor 30 and/or a personnel monitor 37.

As shown in Fig. 4, the system 1 also includes a quality predictor 17, a condition predictor 18, a cost predictor 19, a process predictor 31, a resource predictor 35 and a personnel predictor 36. The aforementioned predictors are each designed as data models, for example, and are preferably based on methods of machine learning, in particular on artificial neural networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or nonlinear regression models with or without regularization, support vector machines with linear, polynomial or other kernel functions, or the like. The system of Fig. 4 further includes a resource planner 20, a sequencing planner 21, a maintenance planner 22, a spare parts planner 23 and a personnel planner 24. The price generator 25 of the system of Fig. 4 is used to determine a minimum price for a product for economical production, a probable price for a product based on the current market situation and/or a highest possible, realistic price for a product. The price generator also takes into account the general costs of production plant 2.

The planning list 26 from FIG. 4 contains information regarding the products to be produced, in particular their quality requirements. 4 shows in particular the flow of information between the individual components of the system 1 for the economic optimization of a process in a production plant 2, in particular a production plant 2 in the metal-producing industry, the non-iron or steel industry or the production of master alloys.

The production plant (production) 2 implements the economically optimized process, for example with the aid of the automation system 4 and using personnel. The planning of the economically optimized process can be updated cyclically.

The quality predictor 17 assumes central functions for the economic optimization of the process in the production plant 2 State predictor 18, the cost predictor 19, the process predictor 31, the resource predictor 35 and the personnel predictor 36, since these create essential predictions required for the optimization task during planning. In the broadest sense, these are models for calculating the development of various variables. In order for these to reflect the actual characteristics of production plant 2 sufficiently well, they must be successively adapted to the data situation. This is done by data collection via the process monitor 13, the status monitor 14, the maintenance monitor 15, the planning monitor 28, the resource monitor 29, the quality monitor 16, the profit monitor 30 and the personnel monitor 37. Through various learning processes, the quality predictor 17, the status predictor 18, the Cost predictor 19, the process predictor 31, the resource predictor 35 and the personnel predictor 36 updated in their behavior. If necessary, entries from the quality catalog 9, the process catalog 10, the maintenance catalog 11, the cost catalog 12, the spare parts catalog 27, the resource catalog 32, the personnel catalog 33 and the profit catalog 34 are used and/or generated/modified by the predictors mentioned.

An exemplary linking of the quality predictor 17, the condition predictor 18, the cost predictor 19, the process predictor 31, the resource predictor 35 and the personnel predictor 36 with the other components of the system 1 from FIG. 4 is shown in FIG. 5 generally shows the linking of catalogs and models with the predictors mentioned.

The links between the quality predictor 17 and the other components of the system 1 are explained in more detail below on the basis of FIG. 5 . The quality predictor 17 uses the historical records of the condition monitor 14, the process monitor 13 and the quality monitor 16. The State monitor 14 indicates, for example, when the production plant 2 was in which classifiable or formally describable state. The process monitor 13 contains information about which type of process has taken place with which parameter values, insofar as this can be detected in the broadest sense by sensors or by target specifications. The quality monitor 16 contains information on quality properties achieved, for example determined from test results, customer feedback, internal quality assessment or the like. Optimization methods, in particular machine learning methods, are used to generate a prediction model which estimates the product quality achieved for a given product, given a given system status and for a predicted process. The links of the quality predictor 17 with the other components of the system 1 described above by way of example can also be seen for the other predictors, in particular the condition predictor 18, the cost predictor 19, the resource predictor 35 and/or the personnel predictor 36, in FIG.

FIG. 6 shows an exemplary linking of components of a system 1 according to the invention for resource planning.

The resource monitor 29 records the current supply of raw materials and the inflows that have already been planned. The current production planning, ie a sequence and a chronological sequence for entries from the planning list 26 , is stored in the planning monitor 28 . Together, this information allows the resource predictor 35 to predict resource inventories at future points in time. As soon as a negative value is forecast here, it becomes apparent to the resource planner 20 that a procurement process must be initiated. He can now propose or even trigger a procurement for all required types of raw materials that are listed in the resource catalog 32, if he has the appropriate authorization for automated trading present. The corresponding costs can be obtained from the cost catalog 12 or, in the event of a discrepancy between the actual procurement costs and the information listed in the cost catalog 12, these can be updated. Based on this information, the sequence planning 21 can then take into account the forecast resource stocks for future production tasks and only schedule products for which there are positive resource stocks at the respective point in time.

The central component in resource planning is the resource planner 20, which takes the data from the resource predictor 35 into account. The resource predictor 35 takes into account the data from the planning monitor 28 and the resource monitor 29.

The data from the resource planner 20, the resource predictor 35 and the resource monitor 29 are linked to the data from the resource catalog 32, which in turn can exchange data with the cost catalog 12.

The sequence planner 21 takes into account the results of the resource planner 20 and control over the automation system 4, in particular the

Plant automation, the production 2 in the industrial plant.

The data from the sequence planner 21 are also taken into account by the planning monitor 28 , which in turn is linked to the maintenance monitor 15 and the planning list 26 .

The data from the cost predictor 19 and the price generator 25 can also be taken into account for business optimization. FIG. 7 shows an exemplary linking of components of a system 1 according to the invention for the procurement of spare parts. When optimizing the procurement of spare parts, the status monitor 14 records the current plant status, which is also represented by possible entries from the process catalog 10, i.e. characteristic states that have a possible influence on product quality or plant performance, being listed. From the maintenance monitor 15 it can be seen which type of maintenance activity is already planned and at what time, insofar as this is listed in the maintenance catalog 11 . From this it can also be seen which process status from the process catalog 10 is affected by a maintenance measure. The third essential piece of information that can be seen from the planning monitor 28 is which products (recorded with their properties in the planning list 26) are scheduled for production at what time. Based on this information, the state predictor 18 can predict the future state of the system and the process predictor 31 can predict characteristic process states. With the latter information, the quality predictor 17 can in turn determine how the achievable product quality will be in the future. From a comparison between the prediction of the quality predictor 17 with the quality requirements of the products according to the current

production planning (the individual quality criteria of the products are listed in Quality Catalog 9), it can then be determined whether quality features cannot be achieved at a future point in time or can only be achieved with a greatly reduced probability. In this case, the maintenance planner 22 plans suitable measures from the maintenance catalog 11 that allow an improvement in the state of the installation or the state of the process, so that the corresponding process states for the determined deviations from the quality criteria are avoided. For the implementation of

Maintenance measures can then be determined on the basis of the information from the maintenance catalog 11, which also contains a list of the tools, replacement parts and replacement parts required for a maintenance measure, whether these are available in sufficient quantity. If this is not the case, a recommendation for the procurement of spare parts can be made be pronounced or, if there is authorization for automated trading, a procurement process can also be triggered directly. The costs can be taken from the catalog of costs 12 or, after procurement has taken place, the entries from the catalog of costs 12 can be updated with the actual procurement costs.

The procurement of spare parts is based on the data from the spare parts planner 23 , which uses the data from the maintenance planner 22 and the spare parts catalog 27 . The result of the spare parts planner 23 is sent to the sequence planner 21 for further consideration when planning the production sequence.

The spare parts catalog 27 takes into account the data from the cost catalog 12 and the maintenance catalog 11, with the maintenance catalog 11 also making its data available to the maintenance planner.

Other essential components in the procurement of spare parts are the maintenance monitor 15, the planning monitor 28 and the process predictor 31. These are based in part on the data from the quality predictor 17, the

condition predictor 18, the planning list 26, the quality catalog 9 and the process catalog 10.

The sequence planner 21 controls the production 2 in the industrial plant by means of the automation system 4. FIG. 8 shows an exemplary linking of components of a system 1 according to the invention for the procurement of spare parts with personnel planning.

In addition to the procurement of spare parts and replacement parts, as explained in detail in connection with the exemplary embodiment from FIG. 7, the necessary personnel planning can also be carried out in parallel. Since the Since maintenance work is generally carried out by personnel, it must be ensured that the required personnel are available at the required time. For this purpose, the current availability of personnel is recorded in the personnel monitor 37, insofar as existing production processes and already planned maintenance measures make use of this. The personnel predictor 36 can then generate the information as to which personnel resource is available at what point in time in the future or has already been scheduled. For a maintenance measure to be planned, the information can then be taken from the maintenance catalog 11, how many people with which qualifications (such as basic automation, hydraulics, mechanics, etc. - this can basically be so specific that for each individual subtask as part of a maintenance measure there is a reference for the corresponding personnel resource) and for which period of time are required. The personnel planner 24 can then compare the respective requirement with the availability and schedule appropriate personnel resources for the required use. The information from the personal monitor 37 is thus adjusted. Subsequent to a maintenance measure that has been carried out, it can also be checked to what extent the information from the maintenance catalog 11 with regard to the required competence and the required duration of participation is correctly displayed and, if necessary, an update can be made.

The central component in personnel planning is the personnel planner 24 which is based on the data from the personnel predictor 36 and the maintenance planner 22 . The personnel predictor 36 receives data from the planning monitor 28 and the personnel monitor 37, which in turn can access the personnel catalog 33. The personnel planner 24 can make the personnel monitor 37 available.

The data from the personnel planner 24 are made available to the sequencing planner 21 , the sequencing planner 21 controlling production 2 in the industrial plant by means of the automation system 4 . Other relevant components for personnel planning are, for example, the planning list 26, the quality catalog 9, the process catalog 10, the maintenance catalog 11, the maintenance monitor 15, the condition predictor 18, the quality predictor 17 and the process predictor 31.

FIG. 9 shows a further exemplary linking of components of a system 1 according to the invention for the economic optimization of a process in a production facility. The exemplary embodiment from FIG. 9 differs from the exemplary embodiment from FIG. The overall planner 38 can take over the functions of one or more of the aforementioned planners. Otherwise, the exemplary embodiment from FIG. 9 corresponds to the exemplary embodiment from FIG.

10 - 13 show exemplary time sequences for the economic optimization of a process in a production plant 2. According to the invention, the process in the production plant 2 is economically optimized in that the data from the production planning system 3, the automation system 4, the condition monitoring system 5 and the maintenance planning system 6 be linked to the data of the corporate planning system 7.

The situation is constantly changing due to the ongoing production process, new customer orders, unforeseeable changes in plant availability, the product quality achieved, etc. In order to always achieve a situationally economically optimal mode of operation, it is therefore necessary to constantly adapt the various operational plans to the changing boundary conditions. 10 shows an exemplary embodiment of an economic optimization of the process in a production plant 2 by an overall optimizer 38, which implements an overall, economically optimized mode of operation through repetitive planning. The operating status in the production plant changes as a result of external circumstances and the operational processes in the production plant 2. The plan is regularly updated on the basis of these changes. This can take place, for example, at specified intervals and/or after specific events. Due to an updated plan, the operational processes in the production plant 2 are adjusted, for example by means of the automation system 4.

The embodiment of FIG. 11 differs from that

Embodiment from Fig. 10 in that several separate planners are used instead of the overall planner 38. The separate planners are, for example, a resource planner 20, a sequence planner 21, a maintenance planner 22, a spare parts planner 23 and/or a personnel planner 24. The separate planners each optimize different aspects of the production process in the production plant 2 from an economic point of view. This is symbolized in FIG. 11 by plant parts I and II.

The external circumstances and the operating processes have different influences on the aspects to be optimized, so that the operating states for the different aspects of the production plant 2 to be optimized change differently due to the different operating processes and external circumstances. The optimization by the separate planners can in turn take place at specified intervals and/or after specific events. The time intervals and/or events for the separate planners can be designed differently.

The embodiment of FIG. 12 differs from that

Exemplary embodiment from FIG. 10 in that predictors are used to analyze the operating states and to make predictions about future operating states. These predictions are then used in the business optimization of the filing process in the

Production plant considered by the overall optimizer 38.

In the exemplary embodiment from FIG. 13, the optimization is carried out, as in the exemplary embodiment from FIG Predictors for separate aspects of the production plant 2 are provided, such as a quality predictor 17, a condition predictor 18, a cost predictor 19, a process predictor 31, a resource predictor 35 and/or a personnel predictor 36.

Otherwise, the exemplary embodiments from Fig. 12 and Fig. 13 correspond to the exemplary embodiments from Fig. 10 and Fig. 11.

Reference List

1 systems

2 production plant 3 production planning system

4 automation system

5 condition monitoring system

6 maintenance planning system

7 Enterprise Planning System 8 Optimizer

9 quality catalogue

10 process catalogue

11 Maintenance Catalogue

12 Catalog of costs 13 Process monitor

14 Health Monitor

15 maintenance monitor

16 quality monitor

17 quality predictor 18 state predictor

19 cost predictor

20 resource planners

21 sequence planner

22 Maintenance planner 23 Spare parts planner

24 staff planner

25 price generator

26 planning list

27 Spare parts catalog 28 Planning monitor

29 Resource Monitor 30 profit monitor

31 process predictor

32 resource catalog

33 Personnel Catalog 34 Profit Catalogue

35 resource predictor

36 Personnel Predictor

37 staff monitor

38 overall planner

Claims

Expectations

1. System (1) for the economic optimization of a process in a production plant (2), in particular a production plant (2) in the metal-producing industry, the non-iron or steel industry or the production of master alloys, the production plant (2) comprising: a production planning system ( 3) for determining a production sequence for the production plant (2), in particular control data for an automation system (4), the automation system (4) for controlling production in the production plant (2), the automation system (4) preferably using the production planning system ( 3) executes a specific production sequence, a status monitoring system (5) for recording the status of the production plant (2) and its components, a maintenance planning system (6) for determining maintenance measures for the production plant (2), a corporate planning system (7 ) to the business Ve Management of production and maintenance in the production plant (2), and a quality predictor (17), a condition predictor (18), a cost predictor (19), a process predictor (31), a resource predictor (35) and/or a personnel predictor (36) , characterized in that the system (1) comprises an optimizer (8) which combines the data from the production planning system (3), the automation system (4), the condition monitoring system (5) and the maintenance planning system (6) with the data from the enterprise planning system (7 ) linked in order to economically optimize the process in the production plant (2), wherein the optimizer (8) uses the predictors (17, 18, 19, 31, 35, 36) to determine the future plant states, product qualities that can be achieved in the future, future costs or similar relevant properties, parameters or the like in the economic optimization of the process in the production plant ( 2) taken into account.

2. System (1) according to claim 1, wherein the system (1) determines an economically optimized production sequence, economically optimized setpoint specifications for the automation system (4), economically optimized maintenance measures and/or economically optimized administrative measures.

3. System (1) according to claim 1 or claim 2, wherein the system (1) creates a personnel deployment plan for production and/or maintenance, a resource procurement plan, an energy procurement plan, a spare parts procurement plan, a maintenance plan or the like as part of the business optimization.

4. System (1) according to one of claims 1 to 3, further comprising a quality catalog (9), a process catalog (10), a maintenance catalog (11), a cost catalog (12), a spare parts catalog (27), a resource catalog (32 ), a personnel catalog (33) and/or a profit catalog (34).

5. System (1) according to any one of claims 1 to 4, further comprising a process monitor (13), a status monitor (14), a maintenance monitor (15), a planning monitor (28), a resource monitor (29), a quality monitor (16 ), a profit monitor (30) and/or a personnel monitor (37).

6. System (1) according to claim 5, wherein the predictions of the quality predictor (17), the state predictor (18), the cost predictor (19), the process predictor (31), the resource predictor (35) and/or the personnel predictor (36) and/or the economic optimization of the optimizer (8).

Basis of the process monitor (13), the condition monitor (14), the maintenance monitor (15), the planning monitor (28), the resource monitor (29), the quality monitor (16), the profit monitor (30) and/or the personnel monitor (37) be continuously adjusted

7. System (1) according to one of claims 1 to 6, wherein the quality predictor (17), the condition predictor (18), the cost predictor (19), the process predictor (31), the resource predictor (35) and/or the personnel predictor ( 36) is designed as a data model.

8. System (1) according to claim 7, wherein the respective data model is based on machine learning methods, in particular on artificial neural networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or non-linear regression models with or without regularization, Support vector machines with linear, polynomial or other kernel functions, or the like.

9. System (1) according to one of claims 6 to 8, wherein the cost predictor (19) is designed as a price catalog in which costs per unit of consumption or per unit of time are specified or from which the costs can be calculated according to a defined formula.

10. System (1) according to any one of claims 1 to 9, further comprising a resource planner (20), a sequence planner (21), a maintenance planner (22), a spare parts planner (23) and/or a personnel planner (24).

11. System (1) according to one of claims 1 to 10, further comprising a price generator (25) for determining a minimum price for a product for economic production, a probable price for a product based on the current market situation and/or a highest possible realistic price for a product.

12. System (1) according to claim 11, wherein the price generator (25) takes general costs of the production plant (2) into account.

The system (1) of claim 11 or claim 12, wherein the price generator (25) comprises a static price catalog and/or a dynamic pricing algorithm.

14. System (1) according to any one of claims 1 to 13, further comprising a planning list (26) comprising information regarding the products to be produced, in particular their quality requirements.

15. System according to any one of claims 1 to 14, wherein the optimizer (8) is based on algorithms of reinforcement learning or operation research, in particular policy networks, neural networks, linear programs, convex optimization, mixed integer optimization, or the like.

16. A method for the economic optimization of a process in a production plant (2), in particular a production plant (2) in the metal-producing industry, the non-iron or steel industry or the production of master alloys, the production plant (2) comprising: a production planning system (3) for Determination of a production sequence for the production plant (2), in particular control data for an automation system (4), the automation system (4) for controlling production in the production plant (2), the automation system (4) preferably determining the production planning system (3). executes the production sequence, a status monitoring system (5) for recording the statuses of the production plant (2) and its components, a maintenance planning system (6) for determining maintenance measures for the production plant (2), a corporate planning system (7) associated with the production plant (2) for business administration ve Management of production and maintenance in the production plant (2), and a quality predictor (17), a condition predictor (18), a cost predictor (19), a process predictor (31), a process predictor (31), a resource predictor (35) and/or or a personnel predictor (36), the method combining the data of the production planning system (3), the automation system (4), the condition monitoring system (5) and the maintenance planning system (6) with the data of the

Business planning system (7) linked in order to optimize the process in the production plant (2) from an economic point of view, the optimizer (8) using the predictors (17, 18, 19, 31, 35, 36) the future plant states, product qualities that can be achieved in the future, future Cost or similar relevant properties, parameters or The like taken into account in the economic optimization of the process in the production plant (2).

17. The method according to claim 16, wherein the method is an economically optimized one

Production order, economically optimized setpoint specifications for the automation system (4), economically optimized maintenance measures and / or economically optimized administrative measures determined

18. The method according to claim 16 or claim 17, wherein the method creates a personnel deployment plan for production and/or maintenance, a resource procurement plan, an energy procurement plan, a spare parts procurement plan, a maintenance plan or the like as part of the business optimization.

19. The method according to any one of claims 16 to 18, further comprising creating and / or managing a

Quality catalog (9), a process catalog (10), a maintenance catalog (11), a cost catalog (12), a spare parts catalogue, one

resource catalog (32), a personnel catalog (33) and/or a profit catalog (34).

20. The method according to any one of claims 16 to 19, further comprising creating and / or managing a

process monitor (13), a condition monitor (14), a maintenance monitor (15), a planning monitor (28), a resource monitor (29), a quality monitor (16), a profit monitor (30) and/or a personnel monitor (37).

21. The method of claim 20, wherein the predictions of the quality predictor (17), the condition predictor (18), the cost predictor (19), the process predictor (31), the resource predictor (35) and/or the personnel predictor (36) and/or the economic optimization of the optimizer (8 ) based on the process monitor (13), the condition monitor (14), the maintenance monitor (15), the planning monitor (28), the resource monitor (29), the quality monitor (16), the profit monitor (30) and/or the personnel monitor ( 37) are continuously adjusted.

22. The method according to any one of claims 16 to 21, wherein the quality predictor (17), the condition predictor (18), the cost predictor (19), the process predictor (31), the resource predictor (35) and / or the personnel predictor (36) respectively based on a data model.

23. The method according to claim 22, wherein the respective data model is based on machine learning methods, in particular on artificial neural networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or non-linear

regression models with or without regularization, support vector machines with linear, polynomial or other kernel functions, or the like.

24. The method according to any one of claims 21 to 23, wherein the cost predictor (19) is designed as a price catalog in which costs per consumption unit or per unit of time are specified or from which the costs can be calculated according to a defined formula.

25. The method according to any one of claims 16 to 24, further comprising creating and/or managing a resource planner (20), a sequence planner (21), a maintenance planner (22), a spare parts planner (23) and/or a personnel planner (24).

26. The method according to any one of claims 16 to 25, further determining a minimum price for a product for economical production, a probable price for a product based on the current market situation and/or a highest possible, realistic price for a product.

27. The method according to claim 26, wherein the minimum price, the probable price and/or the highest possible, realistic price takes general costs of the production plant (2) into account.

28. The method according to claim 26 or claim 27, wherein the determination of the minimum price, the probable price and/or the highest possible, realistic price is based on a static price catalog and/or an algorithm for dynamic

Pricing based.

29. The method according to any one of claims 16 to 28, further comprising the creation and / or management of a planning list (26) comprising information regarding the pending production

Products, in particular their quality requirements.

30. The method according to any one of claims 16 to 29, wherein the method is based on algorithms of reinforcement learning or operation research, in particular policy networks, neural Meshes, linear programs, convex optimization, mixed integer optimization, or the like.

31. A computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to one of claims 18 to 34, in particular comprising instructions which cause the system (1) according to one of claims 1 to 15 carrying out the method according to any one of claims 16 to 30.