The FLEX-NX300 is an Elevator Group Supervisory Control (GSC) System that is well-suited for large-scale buildings.
The FLEX-NX300 GSC System consists of the Intelligence Computer, which incorporates the Neuro GSC, the Fuzzy GSC and the Virtual Passenger Optimization Module, along with two supervisory computers providing high-speed processing and high reliability.
Featuring the Advanced Neuro Group Supervisory Control
This system is an advanced Group Supervisory Control System which automatically learns control rules suited to traffic conditions in various building through a neural network.
The neural network is modeled after human brain functions and is capable of learning the optimal elevator car assignments by itself.
Adoption of the Neuro GSC System will substantially improve the accuracy of the advance notice of approaching elevators.
Flexibly Responding to In-Building Changes
There is no need to modify the program even if the traffic demand fluctuates as a result of a change of tenants or the relocation of tenants to other floors in the building.
The neural network incorporates a self-learning ability that automatically executes updates while keeping elevators in operation, even if an unexpected circumstance arises.
Further Reduction in Passenger Waiting Time
With the adoption of the Virtual Passenger Optimization Method, which probabilistically extrapolates the passenger waiting time including passengers expected to arrive at a stop in the future, the prediction accuracy in the group supervisory control will be enhanced,
thereby further reducing the waiting time.
Reliable Multiple Backup Systems
Normally, the Neuro Supervisory Control is implemented while the Intelligence Computer and the Supervisory Computer Ⅰ are working together.
In addition, as a backup system, the Neuro Supervisory Control can continue operations by working with the Supervisory Computer Ⅱ, should the Supervisory Computer Ⅰ fail.
Neuro Group Supervisory Control
The Neuro Group Supervisory Control System consists of the Fuzzy GSC and the Neuro GSC. The Fuzzy GSC determines which elevator car is assigned to a call by utilizing the knowledge of the group supervisory control expert and the Neuro GSC checks the car assigned by the Fuzzy GSC to determine if there is another more suitable car.
When a hall call is generated, the Fuzzy GSC recommends a suitable elevator car, based on the statistical-analysis learning function, knowledge database, and the waiting-time evaluation index that is extrapolated by the Virtual Passenger Optimization Module, and then the Fuzzy GSC sends the data to the Neuro GSC. In the Neuro GSC, checks are made to the call-assigned cars by using the neural network that acquires knowledge in the same way as the human brain. In this way, the most suitable elevator car is assigned while the two GSCs are complementing and cooperating with each other.
