Scheduling systems and control systems are needed in the paper production
industry to ensure that all plants in the mill operate correctly. An
is used to schedule the paper production machines. Other plants are
scheduled algorithmically according to the paper production schedule.
The Tomakomai Mill consists of ten paper making machines, energy supply plants and pulp supply plants. Two hundred paper products are produced per month. Each product has a specified production volume and due date, and requires a specified machine to produce. There are twenty-five kinds of pulp, and products are made from several mixtures of the pulp. Hence, the mixture ratio characterises the product nature.
In the Tomakomai Mill, a millwide production management system exists. It supports all of the production activities, such as energy supply, pulp supply, paper making and shipment. The system has a planning level and a control/operation level. The scheduling system is situated on the planning level for paper making. It receives product orders from the headquarters office, makes a schedule and delivers it to the other planning systems. Each system schedules and optimises its operations based on the paper making schedule.
Expert System Overview
The paper production scheduling system consists of an
for automated scheduling, and a data management system. The expert system
consists of three subsystems: product group scheduling system, individual
scheduling system, and a balancing scheduling system. The group scheduling
system makes a rough schedule, the individual scheduling system details the
result of the rough schedule, and the balancing schedule system optimises the
result of the individual system.
In practice, it is important to modify the schedule created by the expert system or to reschedule after modification. The scheduling expert system has a schedule editor. Users can add, delete, shift and swap products.
The expert system development required 27 months. The development consisted of three stages: feasibility model, prototype model, and operational model.
In each scheduling subsystem, knowledge is classified into three groups:
task control knowledge, domain knowledge and strategic knowledge
Each scheduling consists of several tasks. Each task is concerned with
domain knowledge and strategic knowledge. Task control knowledge defines
the general flow of the scheduling .
This knowledge arises from problem solving behaviour exhibited by the human
experts. This knowledge is
Domain knowledge represents problem characteristics. In the paper production industry, domain knowledge consists of pulp mixture ratio, energy consumption rate for an individual product and combination constraints. Some of the knowledge is acquired from the product database. For example, the mixture ratio data can be acquired directly from the products database. Other parts of knowledge are acquired from human experts. For example, low quality paper should be produced just after the shut down period. The constraints are represented by frames.
Strategic knowledge affects the scheduling efficiency. In the group product scheduling and individual scheduling, job ordering knowledge is one of the most important kinds of knowledge. This knowledge is represented by evaluation functions. The initial arrangement strategy, which determines the initial place for selected jobs according to job ordering knowledge, is represented by production rules. Backtracking strategies are also represented by production rules.
System Implementation and Evaluation
The scheduling systems are implemented with an expert shell utility, ASIREX.
The main considerations are that:
This scheduling system has been in practical use since January 1989. The greatest advantage of this system is its scheduling speedup. The scheduling time for a monthly schedule was reduced to 2 hours from 3 days, when done by a human expert. Also when human experts scheduled manually, a lack of balance occasionally resulted. Utilising the expert system, all constraints are satisfied. Also, the schedule can be changed whenever necessary; for example, when urgent orders need to be added.