We have developed an energy-saving operations system featuring remote operation of central monitoring equipment installed in a building. This system applies robotic process automation to remote operation to automatically perform energy-saving operations on behalf of the operations manager. Furthermore, as another feature, the system requires only a local area network to connect to the central monitoring equipment enabling automatic operation to be performed regardless of the specifications of the central monitoring equipment. The items targeted for energy-saving operation by this system are the optimal operation of a heat source system, setting of the supply water temperature of heat source equipment, setting of room temperature, and setting of outside-air intake volume. At present, the operations manager has the role of performing these energy-saving operations, but finding the optimal value for each of these operations is a difficult task. An operations manager, moreover, is responsible for tasks other than facility operations such as maintenance management, so changing optimal settings accurately at regular intervals on an ongoing basis can be quite a burden. This system uses robotic process automation technology, so it is capable of performing all energy-saving operations that can be executed by the central monitoring equipment. We installed this system in a large-scale shopping mall and performed energy-saving operations on outside-air processing units. In this trial, we achieved a 44% reduction in the amount of energy required for outside-air processing and a 47% reduction in CO₂ emissions.

The Coefficient of Performance of an ordinary water-cooled chiller is presented as a relationship with the chiller load factor and cooling water temperature. However, the cooling water temperature fluctuates according to the processed heat of the cooling tower originating in the cooling energy of the chiller and to the outside temperature and humidity. It is therefore difficult to obtain the cooling water temperature under the processed-heat and weather conditions at the time of evaluation. This, in turn, makes it difficult to determine the Coefficient of Performance of a water-cooled chiller at the evaluation time. In this research, we formulated the Coefficient of Performance of a water-cooled chiller as a relationship with the chiller load factor and specific enthalpy of outside air. Specifically, we used the Number of Transfer Units (NTU) model of a cooling tower to calculate the cooling water temperature corresponding to the cooling-tower load factor targeting a counterflow cooling tower for a range of values of outside-air specific enthalpy. This technique makes it possible to evaluate the Coefficient of Performance of a water-cooled chiller without determining the cooling water temperature. Furthermore, for the case of installing multiple units of chillers, it becomes possible to calculate the composite Coefficient of Performance of those chillers without having to determine the cooling water temperatures for the different operation load factors of those chillers. Moreover, since the composite Coefficient of Performance can be calculated by combining the different installation capacities of these chillers, the energy consumption of multiple chillers can be calculated at the basic planning stage.

The particle swarm optimization (PSO) concept simulating a simplified social milieu, is optimization useful for solving nonconvex continuous optimization problems. We discuss a new learning algorithm for simultaneously adjusting connection weights in neural networks and user-specified parameters included in units. Based on our algorithm, neural network learning, e.g., learning cost or adaptability, can be improved, as demonstrated in mattress decision support system.