System and method for cooling vehicle computing device
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mass production vehicles with autonomous driving compute clusters face significant heat generation challenges due to high computation demands, making efficient heat removal difficult to maintain high reliability.
Innovation Solution
A vehicle computer cooling system that includes a coolant-based subsystem with a proportional flow valve, switching valve, and multiple compressors to dynamically manage heat dissipation by routing coolant through radiators and condensers, optimizing cooling based on thermal loads and ambient conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a compute cluster is used to execute autonomous vehicle driving computations, then computation capability is improved, but heat generation increases
Solution Approach 1:
The patent extracts the heat removal function from the compute cluster system by introducing a separate, dedicated cooling subsystem. This cooling subsystem includes coolant circulation paths, radiators, and thermal management components that are distinct from the computation hardware, allowing the compute cluster to maintain high computational power while the separate system handles heat dissipation independently.
Solution Approach 2:
The patent introduces coolant as an intermediary substance to transfer heat from the compute cluster components to radiators and heat exchangers. The coolant circulates through the system, absorbing thermal energy from high-power computing devices and transporting it to dedicated cooling components, thereby mediating the heat transfer process and enabling effective thermal management.
2Loss of energy
If cooling strategies are adjusted based on thermal loads, then cooling efficiency is improved, but system complexity increases
Solution Approach 1:
The patent implements dynamic cooling strategies where the cooling subsystem actively adjusts its operation based on real-time thermal loads from the compute cluster. Controllers monitor temperature sensors and modulate coolant flow rates, radiator fan speeds, and compressor operations to match the actual thermal demands, transitioning the system from static to dynamic operation for optimal cooling efficiency.
Solution Approach 2:
The patent incorporates feedback mechanisms through temperature sensors distributed throughout the compute cluster and cooling subsystem. These sensors provide real-time temperature data to controllers that adjust cooling component operations accordingly, creating closed-loop control systems that continuously optimize cooling performance based on actual thermal conditions while managing system complexity through intelligent control algorithms.
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
Effectively manages heat dissipation in autonomous driving compute clusters by anticipating thermal loads and adjusting cooling strategies, ensuring efficient operation and maintaining system reliability.
Implementation Method 1
A coolant-based subsystem with a proportional flow valve, switching valve, and multiple compressors to dynamically manage heat dissipation by routing coolant through radiators and condensers
Implementation Method 2
routing coolant through radiators and condensers, optimizing cooling based on thermal loads and ambient conditions
Data Source
AI summary
A vehicle computer is programmed to determine an expected thermal load for an autonomous driving compute cluster and a target temperature for the compute cluster under the expected thermal load. Based on a comparison of a current temperature of the compute cluster to an ambient temperature, the vehicle computer is further programmed to operate a switching valve in a first coolant path to open the path to one of a radiator portion and a condenser portion and operate, with the path open to the condenser portion, a pump in the first coolant path and a condenser in a second coolant path at speeds respectively based on at least one of the expected thermal load and the target temperature.