By leveraging AI technology and HVAC expertise, the "AI-based
energy-saving optimization system of data center" integrates
multiple aspects such as building automation (BA), power, and
dynamic environment monitoring to achieve precise, dynamic,
on-demand cooling, centralized management, optimal control, and
automatic adjustments for the cooling system. This effectively
reduces the energy consumption of the cooling system, ultimately
lowering the overall PUE index of the data center.
Core Value
All-in-one aggregation:
• Breaks the "information silos" of independent operation of
professional subsystems to form an integrated, full-link energy
management and control data platform, and builds a digital base for
energy efficiency analysis and energy-saving optimization
Refined analysis:
• Multi-dimensional, multi-view energy consumption display and
energy efficiency analysis, refined insights into energy
destination, and identification of energy-saving space based on
energy consumption distribution
Algorithm strategy:
• The integrated AI-based algorithm model conducts systematic,
end-to-end operational diagnosis based on real-time operating
parameters, and develops control strategies to improve energy
efficiency
Safe automatic control:
• The energy-saving strategy is sent to the equipment via
professional systems such as BA and dynamic environment monitoring.
The automatic control of the HVAC system is implemented on the
basis of ensuring the operation safety, while the real-time
optimisation of PUE is achieved
Continuous optimization:
• AI model training and AI data inference are conducted with
accumulated operating data under different operating conditions, to
improve the accuracy and adaptability of the AI-based algorithm
model and enable the continuous optimal operation of the
refrigeration system.
Application Scenarios
Medium to large/super large data centers with (water/air cooled)
chilled water systems.
Functions
• Data management
Data management, consolidation and analysis based on a unified
object model and professional data requirements, to build a
database for energy efficiency management and energy-saving
optimization
• Topology display
Electric and HVAC system topologies are mapped according to the
actual on-site conditions of the data center with the configuration
tools to help present the equipment operation status, key operating
parameters and the information of energy indicators and provide
users with a full-scene and full-link energy board
• Alarm management
Multi-dimensional alarm trigger conditions, including thresholds,
changes, actions and trends are provided to support the setting of
multi-level, multi-threshold alarm and templated batch alarm rules,
helping users to keep up with energy use emergencies in time
• Energy efficiency management
Multi-dimensional energy efficiency monitoring, analysis and
display functions of space, subsystems and equipment are configured
to generate early warning based on trend analysis, support the
accounting of electricity consumption cost, and provide data
support for green energy-saving operation of the data center
• Energy-saving optimization
The prefabricated AI-based energy-saving algorithm model library,
through the management, analysis and diagnosis of real-time
operating data, gains an insight into energy-saving space, generate
optimization strategies, and adopt automatic/manual control modes
to achieve strategy distribution and consequently achieve dynamic
optimisation of the HVAC system of the data center
• Statement management
The user-defined statement styles are supported to help users
analyze energy utilization from multiple dimensions and multiple
perspectives;
• Authority management
The authority management can be implemented for roles and
workgroups to support independent or centralized authorization to
users.
