Group collective

Group Collective is a conventional group control system using up and down landing call buttons. Hall calls are allocated by:

• Estimating the expected travel distance between each hall call and each elevator.
• Allocating the call to the “nearest” elevator.
• Regularly reviewing allocations, so that if one elevator is delayed, another elevator may be reassigned if it is expected to answer the call sooner.

A load bypass feature is included. This is set to 80% by mass. If an elevator is already heavily loaded, it will bypass hall calls to avoid stopping for passengers who are unlikely to be able to board.

For a more detailed discussion of group control, refer to The Elevator Traffic Handbook: Theory & Practice.

Group control algorithms normally have a range of operating modes, as discussed below.

Uppeak Modes

Uppeak modes are used when there is a high flow of passengers travelling from the Home Floor to upper floors, such as during the morning arrival period in an office building.

The two uppeak modes use the same general modelling assumptions. The difference is the Home Floor loading strategy.

Uppeak 1 Mode

In Uppeak 1 mode, idle cars are returned to the Home Floor using a parking call. The doors do not open automatically on arrival.

This strategy is normally the most efficient uppeak strategy.

The reason is that passengers are effectively directed to load one elevator at a time. Once that elevator has loaded and departed, another elevator is ready to take the next passengers. In a typical uppeak situation, passengers arrive at the Home Floor continuously but not fast enough to require several elevators to be loaded at the same time.

This approach encourages fuller loading of one elevator rather than partial loading of several elevators. It can reduce the number of elevators sitting at the Home Floor with doors open, and can help maintain a regular cycle of elevator departures.

In most uppeak design scenarios, Uppeak 1 should be tried first.

Uppeak 2 Mode

In Uppeak 2 mode, idle cars are also returned to the Home Floor using a parking call, but the doors do open automatically on arrival.

This means that more than one elevator may be available for loading at the same time.

This strategy can be more effective in particularly heavy uppeak conditions, especially where passengers are arriving at the Home Floor faster than they can load into a single elevator. In these cases, Uppeak 1 can lead to a queue of elevators at the Home Floor, with only one elevator actively loading passengers while other elevators wait unused.

For example, in a very heavy traffic situation with large groups of arriving passengers, elevator 1 may complete its round trip and return to the Home Floor, ready to load again, before elevators 7 and 8 have had an opportunity to load. In this situation, allowing multiple elevators to open and load simultaneously can make better use of the group.

Uppeak 2 is therefore useful when the limiting factor is not only elevator round trip time, but also the rate at which passengers can physically load into the elevators.

However, Uppeak 2 is generally less efficient than Uppeak 1 in lighter or moderate uppeak traffic. If several elevators are open at the Home Floor when passenger arrivals are not heavy enough to fill them efficiently, passengers may be spread across multiple elevators. This can lead to partial loading, more departures with spare capacity, and less efficient handling overall.

Practical Difference Between Uppeak 1 and Uppeak 2

The main difference is:

• Uppeak 1: brings idle elevators back to the Home Floor but keeps their doors closed until they are selected for loading. This encourages passengers to fill one elevator at a time.
• Uppeak 2: brings idle elevators back to the Home Floor and opens their doors on arrival. This allows multiple elevators to load at the same time.

Uppeak 1 normally performs better when one elevator can load passengers quickly enough to keep up with arrivals.

Uppeak 2 may perform better when passenger arrival rates are so high that loading only one elevator at a time creates a bottleneck.

Down Peak Mode

In Down Peak mode, the served floors above the Home Floor are divided into sectors, equal to the number of elevators.

Elevators are dispatched to the sectors in turn. After serving the down calls in their allocated sector, elevators may also stop for additional hall calls on the way back to the Home Floor.

This can improve performance during periods of high down traffic flow, such as the end of the working day in an office building.

Auto Mode

In most design scenarios, it is preferable to manually select the relevant mode rather than use Auto mode.

Auto mode is best suited to real systems or all-day simulations, where the traffic pattern changes during the day and the control system needs to change mode automatically.

Detection parameters are available via the Advanced settings button.

Uppeak Detection Logic

The Uppeak detection logic operates as follows:

• When a car leaves the Home Floor, the system checks whether the load exceeds the Load Switch Detection Level.
• If the load is above the threshold, the Uppeak Detection Counter increments by 1.
• If the load is below the threshold, the counter decrements.
• When the counter reaches the On level, the Uppeak program starts.
• The Clock Inhibit Timer reduces the counter by 1 at regular intervals, as defined by the Clock Rate.
• When the counter drops to the Clock Inhibit Level, the timer turns off.
• Once the counter reaches the Off level, which is always set at 1, the Uppeak program is disabled.

Down Peak Detection Logic

Down Peak detection follows the same general principle as Uppeak detection, but the load is measured as cars arrive at the Home Floor rather than when they depart.

This allows the system to detect heavy traffic returning to the Home Floor and trigger Down Peak mode accordingly.

Auto Mode Detection Parameters

Default values are provided for Auto Mode detection parameters, but there are no universal recommended settings.

The most appropriate values depend on the building, the traffic pattern, and the design objective. They are usually established by trial and error.

Lower load switch detection levels will generally cause peak mode to activate sooner, making the control system more responsive to changes in traffic conditions. However, overly sensitive settings may cause unnecessary, frequent mode changes.