Air-conditioning system, machine learning apparatus, and machine learning method
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional air-conditioning systems require high workload to optimize device operating capacity post-installation, as they need to measure data under various conditions to construct a model reflecting device characteristics, which is inefficient.
Innovation Solution
An air-conditioning system incorporating an outside air conditioning device and an air conditioning device, along with a machine learning apparatus that acquires state variables and calculates rewards based on energy consumption to optimize operating capacity, using a reinforcement learning process to adjust the operating capacity of both devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a model is constructed by measuring data under various operation conditions after installation, then the model accuracy reflecting device characteristics is improved, but the workload after installation increases
Solution Approach 1:
The patent applies preliminary action by collecting and utilizing manufacturing data (such as device specifications, design parameters, and factory test results) before the air-conditioning system is installed. This allows the model to be pre-constructed with baseline accuracy, reducing the need for extensive post-installation data collection and model refinement work.
2Use of energy by moving object
If the operating capacity of each device is optimized using the model, then energy saving is improved, but the complexity of model construction and data collection increases
Solution Approach 1:
The patent applies universality by creating a multi-functional model that serves both as a design optimization tool and an operational control system. The model integrates device characteristics, environmental conditions, and energy consumption patterns into a unified framework that can be used across different operating scenarios, reducing the need for separate complex models for each function.
Solution Approach 2:
The patent applies parameter changes by utilizing manufacturing data (such as design parameters, component specifications, and factory test results) to dynamically adjust model parameters. This allows the model to adapt to different operating conditions and device configurations without requiring complete reconstruction, thereby reducing construction complexity while maintaining optimization accuracy.
3Loss of energy
If both outside air conditioning device and air conditioning device operating capacities are optimized, then total energy consumption is reduced, but the control system complexity increases
Solution Approach 1:
The patent applies merging by integrating the control of the outside air conditioning device and the air conditioning device into a unified control framework. The model simultaneously optimizes both devices by considering their interrelationships and combined energy consumption, rather than treating them as separate independent systems, thereby reducing overall control complexity while achieving total energy optimization.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces the post-installation workload by constructing an accurate model through reinforcement learning, optimizing energy consumption and operating capacity, thereby enhancing system efficiency.
Implementation Method 1
a heating medium adjuster that adjusts a state of a heating medium flowing through the outside air conditioning unit
Implementation Method 2
a refrigerant adjuster that adjusts a state of refrigerant flowing through each of the multiple indoor units
Implementation Method 3
performs air conditioning of a target space by cooling outside air and supplying the outside air to the target space
Implementation Method 4
performs air conditioning of a target space by heating outside air and supplying the outside air to the target space
Implementation Method 5
performs air conditioning of the target space by cooling inside air and supplying the inside air to the target space
Implementation Method 6
performs air conditioning of the target space by heating inside air and supplying the inside air to the target space
Data Source
AI summary
An air-conditioning system includes an outside air conditioning device, an air conditioning device, and a machine learning apparatus, and includes a state variable acquiring unit configured to acquire state variables, a learning unit configured to perform learning by associating the state variables with at least either an operating capacity of the outside air conditioning device or an operating capacity of the air conditioning device, and a reward calculating unit configured to calculate a reward that correlates with a total of energy consumption of the outside air conditioning device and energy consumption of the air conditioning device. The learning unit performs the learning by using the reward calculated in a period determined in accordance with a time duration until the total of the energy consumption changes after at least either the operating capacity of the outside air conditioning device or the operating capacity of the air conditioning device has changed.