Which creatures are human teachers? Go, go, go.

Biologists have made high-grade silk threads, tear-resistant parachutes and high-strength cables for temporary suspension bridges through the study of spider silk. Ships and submarines are modeled after fish and dolphins.

Rattlesnake missiles are modern weapons developed by scientists imitating the "hot eye" function of snakes and their tongues are arranged with a natural infrared sensing ability like a camera.

The rocket takes off by using the recoil principle of jellyfish and squid.

Researchers have developed many military camouflage equipment for the army by studying the chameleon's color-changing ability.

Scientists studied frog eyes and invented electronic frog eyes.

Termites not only use adhesive to build their anthills, but also spray adhesive on their enemies through small tubes in their heads. So people made a working weapon-dry rubber shell according to the same principle.

The US Air Force has developed a miniature thermal sensor through the "hot eye" function of poisonous snakes.

Drawing lessons from the bionics principle and the fur structure of terrestrial animals, Chinese textile scientists have designed a small barrel warm fabric with the functions of wind resistance and moisture conduction.

According to the principle that the rattlesnake's buccal fossa can feel the temperature change of 0.00 1℃, human beings invented the tracking and chasing rattlesnake missile. Humans also used the principle of leapfrog to design a toad rammer. Humans imitate the highly sensitive sense of smell of police dogs and create "electronic police dogs" for investigation. Scientists have made the first batch of gas masks in the world according to the unique drug detection ability of wild boar nose.

Bionics is a method that human beings have been using, such as "dolphin skin swimsuit" which imitates the structure of dolphin skin. When scientists were studying whale skin, they found grooves and sinks on it, so a scientist imitated the structure of whale skin and covered the surface of the plane with a film. According to experiments, it can save 3% energy. If planes all over the country are covered with such surfaces, billions can be saved every year. For another example, some scientists have studied spiders and found that spiders have no muscles on their legs. Animals with feet can walk mainly by muscle contraction. Why do spiders walk without muscles? After research, spiders walk not by muscle contraction, but by "hydraulic" structure, so people invented hydraulic walking machines ... in short, they were inspired by nature and invented by imitating their structures. This is bionics. This is one aspect that we learn from nature.

On the other hand, we can also use its principles to get inspiration from natural laws and design (including design algorithms). This is the concept of intelligent computing.

Intelligent computing

Intelligent computing, also called "soft computing" by some people, borrows inspiration from natural laws (biology) and imitates and designs algorithms to solve problems according to its principles. Such as artificial neural network technology, genetic algorithm, evolutionary programming, fire simulation technology and swarm intelligence technology.

Swarm intelligence

Social insects rely on collective strength to feed, defend against enemies and build nests. The "intelligence" shown by this group is called group intelligence. For example, bees collect honey, build nests, ants feed, build nests and so on. Inspired by the cooperation between social insects, we studied this principle and designed a new algorithm to solve this problem based on this principle.

Ant algorithm

When an ant forages, it leaves pheromones along the way. These pheromones, like road signs, left a path sign for later "ants". The ants behind will walk along the path with pheromones (the more pheromones, the stronger the ability to lure ants). Scientists have done experiments on this: draw a path on paper with artificial pheromones and do experiments with ants. As a result, ants walk along the pheromone-coated path.

B

D

Mayidong a

C food

When ants are looking for food, they start from the nest, either along AC or ABC (see above). When they find food, they will return to the nest along the original road and leave pheromones on the road, so when they return along AC, they will leave pheromones on AC twice. And what returns along ABC only returns to point D, because its path is long, so the pheromone is only left once in the AD segment (that is, the pheromone concentration on it is lighter than that on AC), so the foraging ants who leave the nest will walk along the high-concentration AC path at this time ... Finally, most ants will look for food along a shorter path. Using this principle, scientists designed an ant algorithm (finding the shortest path).

The above is a simple principle. Of course, it is necessary to design feasible algorithms to make the model more accurate, such as considering the volatilization of pheromones (that is, the concentration of hormones will gradually decrease with time). ).

Finding the shortest path with ant algorithm

1. A group of ants randomly start from the starting point, meet food, grab food and return along the original road.

2. Ants leave pheromone marks on their way back and forth.

3. Pheromones will gradually evaporate with time (generally can be described by a negative exponential function, that is, multiplied by the factor e-at).

4. The probability of ants choosing a path from the nest is directly proportional to the pheromone concentration on each path.

Ant algorithm can also be applied to many practical problems, such as rebuilding communication routes, managing the company's telephone network, charging users, task allocation and so on.

Don't stop, keep thinking.

Furthermore, each ant is regarded as a neuron, and the communication between them is regarded as the connection between neurons, but the connection at this time is not fixed, but random. That is, a randomly connected neural network is used to describe a group. The essence of this neural network is group intelligence.

Scientists got inspiration from a thick point at the end of dragonfly wings, which is slightly larger than the surrounding area, thus solving the phenomenon that airplane wings are broken due to violent shaking.