Sitting in an air-conditioned room, with a window open and the heater on high, is all in a day%s work for TU PhD Bing Yu. He designed a fault detection and diagnosis system to prevent buildings from consuming so much energy.
Today, buildings account for one-third of the world%s total energy consumption.
,,Not such a big city,” says Bing Yu, PhD student from the Energy Technology group (OCP). He’s referring to Wuhan, the city in central China where he was born. Wuhan has 8 million inhabitants. Delft surely must’ve seemed a pretty different world when Yu came here in 1999 to start his second PhD thesis. His first thesis on refrigeration was completed in 1996 at Shanghai Jiaotong University.
People often comment on Yu’s decision to pursue two PhD degrees. ,,Some people think I’m doing similar research, but that’s not the case. In Shanghai, I analyzed the performance of a refrigeration system. In Delft, I’m researching the best ways to distribute cold or hot air and use it properly.”
Last week, Yu defended his second thesis, for which he developed a method to diagnose indoor climate installations. The impetus to start this research was the fact that, according to the U.S. Department of Energy, in 1999 the energy consumption of buildings accounted for one-third of total energy consumption in the Western world. ,,In Europe, it’s even 40 percent of all energy consumption, with the rest consumed by transport, industry, etc,” Yu says. And this won’t decrease in the near future, because nowadays everybody expects buildings to support air-conditioning systems, powerful computers and, preferably, have large rooms.
Yu asserts that part of the energy waste is due to inefficient and improper operations of the heating, ventilation and air-conditioning systems. Yu: ,,For example, in the winter radiators are too hot, so people may open a window for fresh air, instead of turning the heater off. And when they go home, they forget to turn down the heating.”
Yu cites another example, of defective control valves that cause heaters to be stuck in one position. People might think they’ve turned off the heater, when in fact they haven’t. ,,Generally, it takes a long time before people call building maintenance,” Yu says. ,,In summer, if the air-conditioning doesn%t work, no energy is wasted but people can’t work efficiently because they’re uncomfortable. Fault detection and diagnosis is therefore vital for systems that control indoor climate.”
Yu designed a new approach called Level Orientated Fault Detection and Diagnosis, which analyses the performances of building and installations at different levels. First, the entire building’s energy consumption is analyzed. Faults are detected with the help of a fuzzy neural network, which, modeled on our brain nerves, contains many units (neurons) and processes complicated information. Yu tested this network in an office in Rijswijk during winter. By leaving a window open, Yu introduced faults into thebuilding. When a fault is detected on the first level, the lower level is activated and the rooms are analyzed. For this, Yu developed a model which he studied in an air-conditioned room in which a window was kept open and the radiator valve was jammed with the heat on.
After detection on this second level, further diagnosis is done on the next level, where detailed components are analyzed. For this Yu developed a program that simulates the real-time behavior of indoor climate installations and continuously compares the results of the simulations with the measurements, so mistakes can be diagnosed. To test this last model Yu designed a test facility in a laboratory at the Energy Technology department.
,,It’s like visiting the doctor. A doctor doesn’t immediately x-ray your heart. First, he asks general questions about your health. My approach to analyzing the installations also starts on a general level, before going into the details.”
During his thesis, Yu was inspired by the Chinese saying, ‘Reduce big troubles to smaller ones and reduce these smaller troubles to nothing’. This saying is printed in Chinese across the front cover of his thesis, the characters getting smaller and smaller from top to bottom.
Reducing to nothing didn’t work out completely, however. ,,There’s one disadvantage,” Yu says. ,,Human beings learn from their faults, but installations do not. I only have a small database, or library, of faults which my method can detect. I need the cooperation of industry to increase my library, because some mistakes occur frequently and others perhaps just once every ten years. A library full of knowledge would make the method more applicable. My research was limited to the faults I studied.”
,,We only gave the conceptual design of the fault detection and diagnosis for heating, air-conditioning and ventilation systems,” says Professor Dolf van Paassen, Yu’s promoter. ,,How to apply it in practice is up to the users. We don’t make a final product.” Yu intends to stay on for a while at TU Delft and work to increase the size of his library and to hopefully find building engineers who will test his method. Yu: ,,I do believe my method can help save the energy that%s now being wasted in buildings. But I can’t estimate exactly how much savings.”
Within a year, Yu will return to China. ,,There are more future career opportunities for me in China than in the other countries. In China, the economy%s growing very fast. I got good and useful experience in Delft, met many people and my network is much bigger now.”
Sitting in an air-conditioned room, with a window open and the heater on high, is all in a day%s work for TU PhD Bing Yu. He designed a fault detection and diagnosis system to prevent buildings from consuming so much energy. Today, buildings account for one-third of the world%s total energy consumption.
,,Not such a big city,” says Bing Yu, PhD student from the Energy Technology group (OCP). He’s referring to Wuhan, the city in central China where he was born. Wuhan has 8 million inhabitants. Delft surely must’ve seemed a pretty different world when Yu came here in 1999 to start his second PhD thesis. His first thesis on refrigeration was completed in 1996 at Shanghai Jiaotong University.
People often comment on Yu’s decision to pursue two PhD degrees. ,,Some people think I’m doing similar research, but that’s not the case. In Shanghai, I analyzed the performance of a refrigeration system. In Delft, I’m researching the best ways to distribute cold or hot air and use it properly.”
Last week, Yu defended his second thesis, for which he developed a method to diagnose indoor climate installations. The impetus to start this research was the fact that, according to the U.S. Department of Energy, in 1999 the energy consumption of buildings accounted for one-third of total energy consumption in the Western world. ,,In Europe, it’s even 40 percent of all energy consumption, with the rest consumed by transport, industry, etc,” Yu says. And this won’t decrease in the near future, because nowadays everybody expects buildings to support air-conditioning systems, powerful computers and, preferably, have large rooms.
Yu asserts that part of the energy waste is due to inefficient and improper operations of the heating, ventilation and air-conditioning systems. Yu: ,,For example, in the winter radiators are too hot, so people may open a window for fresh air, instead of turning the heater off. And when they go home, they forget to turn down the heating.”
Yu cites another example, of defective control valves that cause heaters to be stuck in one position. People might think they’ve turned off the heater, when in fact they haven’t. ,,Generally, it takes a long time before people call building maintenance,” Yu says. ,,In summer, if the air-conditioning doesn%t work, no energy is wasted but people can’t work efficiently because they’re uncomfortable. Fault detection and diagnosis is therefore vital for systems that control indoor climate.”
Yu designed a new approach called Level Orientated Fault Detection and Diagnosis, which analyses the performances of building and installations at different levels. First, the entire building’s energy consumption is analyzed. Faults are detected with the help of a fuzzy neural network, which, modeled on our brain nerves, contains many units (neurons) and processes complicated information. Yu tested this network in an office in Rijswijk during winter. By leaving a window open, Yu introduced faults into thebuilding. When a fault is detected on the first level, the lower level is activated and the rooms are analyzed. For this, Yu developed a model which he studied in an air-conditioned room in which a window was kept open and the radiator valve was jammed with the heat on.
After detection on this second level, further diagnosis is done on the next level, where detailed components are analyzed. For this Yu developed a program that simulates the real-time behavior of indoor climate installations and continuously compares the results of the simulations with the measurements, so mistakes can be diagnosed. To test this last model Yu designed a test facility in a laboratory at the Energy Technology department.
,,It’s like visiting the doctor. A doctor doesn’t immediately x-ray your heart. First, he asks general questions about your health. My approach to analyzing the installations also starts on a general level, before going into the details.”
During his thesis, Yu was inspired by the Chinese saying, ‘Reduce big troubles to smaller ones and reduce these smaller troubles to nothing’. This saying is printed in Chinese across the front cover of his thesis, the characters getting smaller and smaller from top to bottom.
Reducing to nothing didn’t work out completely, however. ,,There’s one disadvantage,” Yu says. ,,Human beings learn from their faults, but installations do not. I only have a small database, or library, of faults which my method can detect. I need the cooperation of industry to increase my library, because some mistakes occur frequently and others perhaps just once every ten years. A library full of knowledge would make the method more applicable. My research was limited to the faults I studied.”
,,We only gave the conceptual design of the fault detection and diagnosis for heating, air-conditioning and ventilation systems,” says Professor Dolf van Paassen, Yu’s promoter. ,,How to apply it in practice is up to the users. We don’t make a final product.” Yu intends to stay on for a while at TU Delft and work to increase the size of his library and to hopefully find building engineers who will test his method. Yu: ,,I do believe my method can help save the energy that%s now being wasted in buildings. But I can’t estimate exactly how much savings.”
Within a year, Yu will return to China. ,,There are more future career opportunities for me in China than in the other countries. In China, the economy%s growing very fast. I got good and useful experience in Delft, met many people and my network is much bigger now.”
Comments are closed.