Improvement in Observation State aggregation logic

TMC implements obsstate aggregation to determine effective observation state of the subarray. The existing algorithm works for now but it is primarily based on the observation state events received from lower level devices. This way primarily caters to the needs of aggregation under normal scenarios. In order to handle the complex scenarios arised from  implementation of command timeouts, error handling mechanisms and long running commands it is essential to refactor and update the obsstate aggregation algorithm.

Another drawback is the algorithm blocks the Tango bus for longer duration which lead to inefficient event processing in the Subarray Node.

The refactored algorithm improves the way obsstate events are handled and processed which would improve the efficiency and performance, eliminates the occurrances of race conditions in the code and efficient event processing for bigger scale.

The improved observation state aggregation design is finalized at :  https://confluence.skatelescope.org/pages/viewpage.action?pageId=284165062

The Subarray Node for Mid and Low implements the new algorithm for obsstate aggregation as per the design.

The new TMC Low version 0.18.0 and TMC Mid version 0.22.2 are released on CAR.