Information processing device and air-conditioning system provided with same
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing air-conditioning systems with multiple indoor units face inefficiencies in power consumption due to lack of consideration for individual unit heat loads, leading to increased maximum power usage.
Innovation Solution
An information processing apparatus that communicates with an air-conditioning management system to estimate thermal loads for each unit using learning data and models, determining optimal operating schedules for compressors to achieve set temperatures efficiently across multiple indoor units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single control strategy is applied to multiple indoor units without considering individual heat loads, then device complexity is reduced, but power consumption increases due to inefficient operation
Solution Approach 1:
The patent segments the control system by creating individual learning models for each indoor unit based on its specific heat load characteristics. The control apparatus divides the plurality of indoor units into different control groups, allowing each segment to be optimized independently rather than applying a uniform control strategy to all units.
Solution Approach 2:
The patent implements local quality by tailoring the control strategy to each indoor unit's specific environment and heat load characteristics. Learning models are trained individually for each unit using its own operational data, enabling localized optimization of operating parameters such as compressor capacity and fan speed according to local conditions.
2Temperature
If precooling or preheating operations are performed without accurate heat load forecasting, then target temperature can be reached at target time, but energy efficiency deteriorates due to excessive power consumption
Solution Approach 1:
The patent performs preliminary actions by conducting precooling or preheating operations in advance of the target time, but the duration and intensity of these preliminary actions are precisely controlled based on heat load forecasts generated by trained learning models. This ensures that the target temperature is reached exactly at the target time without excessive energy consumption.
Solution Approach 2:
The patent implements feedback mechanisms where the learning models are continuously trained using actual operational data from indoor units, including temperature changes, power consumption, and environmental conditions. This feedback loop improves the accuracy of heat load forecasts over time, enabling more efficient precooling and preheating operations.
3Use of energy by moving object
If individualized control strategies are implemented for each indoor unit based on learning models, then power consumption is reduced, but device complexity increases
Solution Approach 1:
The patent achieves universality by using a single control apparatus that can manage multiple indoor units with different heat load characteristics. The control apparatus performs multiple functions including data collection, learning model training, heat load forecasting, and real-time control optimization, all within one integrated system that serves the entire air conditioning network.
4Stability of the object's composition
If heat load forecasting accuracy is improved through learning and memorization, then temperature stability is enhanced, but computational requirements and processing time increase
Solution Approach 1:
The patent performs preliminary action by training the learning models in advance using historical operational data, so that when real-time control is needed, the models are already prepared and can provide quick heat load forecasts. This pre-training approach separates the computationally intensive learning phase from the real-time control phase, reducing processing time during actual operation.
Data Source
Figure 1~2
Figure 3~5
Figure 6
AI summary
An information processing apparatus incudes a schedule determination unit configured to determine a schedule of an operating status for causing a temperature of an air-conditioned space of each of a plurality of load-side units to reach a specified set temperature at a specified set time, based on a learning model representing an input-output relationship between input data indicating an influencing factor of a thermal load of each of the load-side units and output data indicating the thermal load.