Flight safety monitoring device for aircraft with alarm

• US 5,710,559 A
• Filed: 03/10/1995
• Issued: 01/20/1998
• Est. Priority Date: 03/28/1994
• Status: Expired due to Fees

• Current Assignee: Bodenseewerk Gerätetechnik Gmbh
• Original Assignee: Bodenseewerk Gerätetechnik Gmbh
• Inventors: Krogmann, Uwe
• Primary Examiner(s): SWARTHOUT, BRENT
• Application Number:
  US08/401,960
• Time in Patent Office: 1,047 Days
• Field of Search: 340/945, 340/963, 340/961, 340/970, 340/439, 364/424.06, 364/439, 395/21, 395/22, 395/23
• US Class Current: 340/963
• CPC Class Codes: G05D 1/0055 with safety arrangements
• CPC Class Codes: G06N 3/042 Knowledge-based neural networks; Logical representations of neural networks
• CPC Class Codes: G09B 9/006 for locating or ranging of objects
• CPC Class Codes: G09B 9/24 including display or recording of simulated flight path
• CPC Class Codes: G09B 9/44 providing simulation in a real aircraft flying through the atmosphere without restriction of its path
• CPC Class Codes: Y10S 706/90 Fuzzy logic
• CPC Class Codes: Y10S 706/913 Vehicle or aerospace

Abstract

For monitoring flight safety of aircraft, a monitoring device accommodated in a pod is provided. The monitoring device has a monitoring unit to be attached to an aircraft. Sensors for measuring aircraft data independently of the sensors of the aircraft itself are provided in the monitoring unit. There is a device for automatically monitoring the operation of the technical equipment of the monitoring unit. In addition, the monitoring unit contains a system for automatically monitoring the pilot'"'"'s reactions on the basis of aircraft data provided by the independent sensors. Furthermore, the monitoring unit contains a device which responds to deviations of the aircraft from the range of safe flight states. This device triggers an alarm. There is a device responding to the aircraft inadmissibly approaching ground and a collision warning device, which responds to the risk of collision with other aircraft. Also these devices trigger an alarm in the case of danger.

Claims

1. A monitoring device for monitoring flight safety of an aircraft containing sensors for aircraft data, comprising:

(a) a monitoring unit (24) adapted to be attached to the aircraft and containing technical equipment;

(b) said technical equipment including sensor means (26,28,30) for measuring said aircraft data independently of said aircraft sensors;

(c) means (36,38) for automatically monitoring the operability of said technical equipment of said monitoring unit (24);

(d) means (36,38) for automatically monitoring the pilot'"'"'s reactions (64) on the basis of said aircraft data provided by said independent sensor means (26,28,30), said reaction monitoring means (36,38) being also mounted in said monitoring unit; and

(e) means for evaluating said monitored reactions to form a measure of the pilot'"'"'s flying ability thereby indicating whether monitored reactions to particular sensed aircraft data conformed with expected reactions.

2. A monitoring device as claimed in claim 1, wherein said monitoring unit (24) contains warning means, to which aircraft data from said independent sensors are applied and which is designed to respond to deviations from safe flight states of the aircraft by providing a warning signal.

3. A monitoring device as claimed in claim 1, wherein said monitoring unit (24) contains ground contact warning means to which aircraft data from said independent sensors are applied and which is designed to respond to inadmissibly close approach of the aircraft to the ground by providing a warning signal.

4. A monitoring device as claimed in claim 1, wherein said monitoring unit (24) contains collision warning means which is designed to respond to the danger of collision with other aircraft by providing a warning signal.

5. A monitoring device as claimed in claim 1, wherein said equipment monitoring means comprises an associative neural network (36) having inputs and outputs, and a rule-based expert system (38) connected to said neural network outputs, said expert system (38) being designed to provide, as output information thereof, information about the flight safety of the aircraft and a decision about measures to be taken.

6. A monitoring device as claimed in claim 5, wherein said rule-based expert system (38) is an associative memory operating with fuzzy logic.

7. A monitoring device as claimed in claim 1, wherein said means (64) for monitoring the pilot'"'"'s reactions comprise:

(a) a neural network with weights which

in an initial state has been trained to model the behavior of a generic pilot as reaction to sensor signals, and

is arranged, in operation, to receive sensor signals indicative of the state of the aircraft from said independent sensor means and signals which represent the reactions of a real, human pilot, whereby the neural network is re-trained during a training process from its initial state to a state with changed weights, whereby its behavior approximates the behavior of the real human pilot,

(b) an adaptive knowledge and data base containing rules which

stores in a memory rules codifying, in an if-then form, the behavior of a pilot as response to sensor signals from said independent sensor means,

said rules being provided with weights which represent the importance of the rules for the pilot'"'"'s behavior,

(c) means for applying said sensor signals from said independent sensor means to said knowledge and data base,

(d) means for changing the weights of the knowledge and data base step-by-step depending on changes of the weights of the neural network relative to its initial state, and

(e) evaluating means for forming a measure of the pilot'"'"'s flying ability from the changed weights of the rules in the knowledge and data base.

8. A monitoring device as claimed in claim 7, wherein said adaptive knowledge and data base is an associative memory operating with fuzzy logic.

9. A monitoring device as claimed in claim 7, and further comprising a computer model (84) of the aircraft dynamics, commands from the pilot being applied both to the aircraft and to said model, said computer model (84) providing estimated state variables of the aircraft;

said estimated state variables of the aircraft being also applied to said evaluating means.

10. A monitoring device as claimed in claim 9 wherein said computer model (84) is a neural network which has been trained on the basis of commands given by the pilot and measured state variables of the aircraft.

11. A monitoring device as claimed in claim 1, wherein said monitoring unit (24) and said independent sensor means (26,28,30) are accommodated in a pod (25) adapted to be attached to the aircraft (56).