Researchers develop a drone that swoops and lands like a bird

研究者发明了一种像鸟儿一样着陆的无人机

One day they may be able to help with fires or deliver packages.

也许有一天它们会被用于救火和运送包裹。

Watch out, birds. The drones are coming for your jobs. Researchers at BMT Defence Services (BMT) and the University of Bristol in Britain have built a fixed-wing UAV that can land as well as its avian counterparts, reports Popular Mechanics. Although BMT's project is currently part of a wider defense program called Autonomous Systems Underpinning Research, the team believes their drone could one day be used for other tasks like putting out fires or delivering packages.

以后你要小心天上的鸟儿咯，它们可能是过来抢你工作的无人机呢！根据《大众机械》的报道，BMT国防服务部和英国布里斯托大学的研究员已经造出了一种像鸟儿一样着陆的固定翼无人机。尽管该项目只是一个大国防项目“自治系统基础研究”中的一小部分，但是该团队相信他们的无人机将来可能会被用来执行灭火或者送快递的任务。

When birds land, they perform a "deep stall," meaning they swoop in at low altitude and angle their wings upward before landing. BMT's drone does this too, thanks to a new morphing wing that can sweep forwards and backwards to create a pitching moment, or twist to allow the aircraft to roll. With this kind of high maneuverability, researchers envision a future where UAVs can easily fly through urban environments, dodging lampposts and power lines.

鸟儿降落时，会做一个“雀降”动作，以一个低高度俯冲下来，然后翅膀向上展开，最后直立缓缓落地。BMT的无人机也这么做，这得益于一种新的可以向前向后伸展的变形机翼，这可以使无人机更倾斜地盘旋。有了这种超高机动性，研发者们可以预见未来无人机可以轻松飞越城市环境，避开灯柱和电线。

But, it takes more than high-tech wings for a drone to safely do a deep stall. The team also had to build it a bird-like brain, one that could compensate for slight changes in speed, wind, angle and wing position. They achieved this through something called "Q-learning," a technique where an artificial intelligence learns an optimal course of action by raising its "Q," or satisfaction level. Like a kid in an arcade, it kept trying to beat its high score, until it worked out how to get from its starting point to its destination. After about 5,000 practice attempts, BMT says the drone pulled off a soft landing without a runway.

然而，无人机要安全地完成“雀降”动作，需要的不仅仅是高科技的翅膀。团队也需要构造一个和鸟一样的大脑，可以补偿微小的速度、风速、角度和机翼位置变化。他们通过一种叫“Q学习”的技术达到了这一目标。此技术就是用人工智能技术学习最佳着陆动作从而提高满意度（Q）。就像游乐场里的小孩会不断试着去打高分，“Q学习”也会不断去冲击更高的分数。经过大约5000次尝试，BMT说无人机终于完成了一次不需要跑道的软着陆。

来源：大众机械/Popular Mechanics